View Full Version : Rejection Rates for tested tier 1 AR15 parts
Does anybody have any hard information on the rejection rates of individually tested AR15 parts such as bolts, carriers, barrels? Obviously all tested parts don't pass, otherwise there would be no reason to test at all. The rejection rate could be useful information.
It is often suggested to purchase only individually tested parts rather than batch tested parts. Common sense suggests you only know the part is good if it is tested. Common sense also suggests that because a part isn't individually tested that it cannot be said it will not pass the tests.
Therefore, I believe it would advantageous for M4C members to know the actual rejection rate of tested parts to be able to suggest the likelihood that non-tested parts don't meet spec and are prone to failure. This information can also be used for newcomers to determine whether or not it is wroth the money to purchase tested parts to meet their needs depending on their individual risk-adverseness.
MarshallDodge
07-20-09, 12:17
I cannot answer your original question but I would like to address your concern about risk.
Purchasing a firearm that has a poor track record typically means that the manufacturer has poor quality/process control and may cut corners. Most of the time these corners are cut in areas that you would not think to look such as small parts, springs, etc. One example of this is the allen screws on the gas key. Here are a couple screws that quality versions cost 15 cents and they are installing cheaper 10 cent versions.
There are a couple AR manufacturers that are way overpriced for the quality that you are getting. Sure you may be saving a couple hundred less than a Tier 1 gun but in reality you are still paying too much.
Absolutely. I'd merely like to see empirical evidence to show people with numbers rather than the abstract of "tested" vs "not-tested".
dookie1481
07-20-09, 13:01
Absolutely. I'd merely like to see empirical evidence to show people with numbers rather than the abstract of "tested" vs "not-tested".
I doubt you ever will. I imagine that info is pretty closely guarded.
Jay
Does anybody have any hard information on the rejection rates of individually tested AR15 parts such as bolts, carriers, barrels? Obviously all tested parts don't pass, otherwise there would be no reason to test at all. The rejection rate could be useful information.
It is often suggested to purchase only individually tested parts rather than batch tested parts. Common sense suggests you only know the part is good if it is tested. Common sense also suggests that because a part isn't individually tested that it cannot be said it will not pass the tests.
Therefore, I believe it would advantageous for M4C members to know the actual rejection rate of tested parts to be able to suggest the likelihood that non-tested parts don't meet spec and are prone to failure. This information can also be used for newcomers to determine whether or not it is wroth the money to purchase tested parts to meet their needs depending on their individual risk-adverseness.
Most companies do not share their rejection data with the consumer. I do know that companies reject parts all the time (especially if they have a .Gov inspector examing parts).
C4
That is very disheartening because it underscores the importance of having good, tested parts vs batch-tested parts.
That is very disheartening because it underscores the importance of having good, tested parts vs batch-tested parts.
I am not sure why you think that some companies are not testing each and every part (no one has said that they do not).
What I did tell you is that companies do not devulge their rejected parts numbers.
C4
The other issue is that you missed is this is apples versus oranges.
Your expecting a Bolt from A that is Individually tested to be the same as Bolt from B that is Batch tested.
The problem is they are not, and thus there is no way to draw a comparible.
Tier1 Manufacturers generally are required to Individual Test if your accepting a Tier 1 Manufacturer of one who has met Schedule and delivers product to the Gov, and thus your further complicating the issue by adding a Tier 2 or Tier 32 manufacturer.
I am not sure why you think that some companies are not testing each and every part (no one has said that they do not).
What I did tell you is that companies do not devulge their rejected parts numbers.
C4
I think you're misunderstanding me.
For instance, it is stressed for bolts to be MPI tested and one should purchase bolts that have had this done over one that has not.
We know all bolts that have been tested and passed are good to go. We know some companies do not test theirs at all.
A bolt that hasn't been tested may or may not pass the test--we don't know.
If we knew the rejection rate of tested bolts, it could be used to demonstrate the probability of non-tested bolts would be a reject by a tier one company. This could be used to provide empirical evidence of why (and how much) better tier one carbines are than tier two and demonstrate the actual level of importance of testing for a consumer rifle.
ETA:
For instance, if we knew that LMT has 50% of their bolts fail MPI testing, we could then suggest to new gun buyers that there is a strong probability that 50% or more of Bushmaster's bolts are of reject quality since they are not tested.
I think you're misunderstanding me.
For instance, it is stressed for bolts to be MPI tested and one should purchase bolts that have had this done over one that has not.
We know all bolts that have been tested and passed are good to go. We know some companies do not test theirs at all.
A bolt that hasn't been tested may or may not pass the test--we don't know.
If we knew the rejection rate of tested bolts, it could be used to demonstrate the probability of non-tested bolts would be a reject by a tier one company. This could be used to provide empirical evidence of why (and how much) better tier one carbines are than tier two.
No, I understand what you are are saying.
Also understand that MP testing bolts that have not first been HPT'd will most likely never show a failure (and never be caught).
The other part of the puzzle is the exit criteria established by the manufacturer. Do they have a zero acceptance policy for inclusions found in the bolt? Or do they allow small ones? How many?
C4
I was using MPI tests generically and any other test can be used interchangeably.
This is very useful information and could be quite useful when helping one decide what carbine to purchase. Thresholds are very important and this could even be useful for comparing tier one companies. You bring up a great point: if we don't know what an individual company will accept, all simply suggesting they test it is very moot. Thanks for the input.
I was using MPI tests generically and any other test can be used interchangeably.
This is very useful information and could be quite useful when helping one decide what carbine to purchase. Thresholds are very important and this could even be useful for comparing tier one companies. Thanks for the input.
Oh sure, but realize the following:
1. Manufacturers are not going to share failure rates.
2. Most "manufacturers" are nothing more than parts assemblers and do not ACTUALLY make anything. So they have ZERO idea what the actual manufacturer has done as far as testing.
3. Most "manufacturers" really do not care if anything has been tested or not. All they care about is the "finish" and cost.
C4
Oh sure, but realize the following:
1. Manufacturers are not going to share failure rates.
2. Most "manufacturers" are nothing more than parts assemblers and do not ACTUALLY make anything. So they have ZERO idea what the actual manufacturer has done as far as testing.
3. Most "manufacturers" really do not care with anything has been tested or now. All they care about is the "finish" and cost.
C4
So what you're suggesting is that we have to rely on reputation and that all parts testing is moot if we do not know the acceptance/rejection threshold? That just because a part is tested (and stamped accordingly) we don't know that it is as good as or better than another brand that is also tested?
So what you're suggesting is that we have to rely on reputation and that all parts testing is moot if we do not know the acceptance/rejection threshold? That just because a part is tested (and stamped accordingly) we don't know that it is as good as or better than another brand that is also tested?
Basically, yes. I would rely on people that work in this industry to be able to spot quality from crap.
I talk to instructors to see which brands of AR's go down the most (as these guys see tons of guns). This is another good way to tell.
C4
This is good information.
I suppose I won't find good organized empirical evidence to suggest brand X over brand Y.
My personal carbines are still awaiting their first failure. Of the three I mostly shoot, I have a 6920 and two frankenguns made of CMMG parts (with F/A BCGs with black extractor inserts)with Wilson Combat bbls. My current build is going to be mostly BCM parts. I've also considered purchasing an additional 6920 or a DD XV and this leads to me to believe there is no good way to compare the two.
This is good information.
I suppose I won't find good organized empirical evidence to suggest brand X over brand Y.
My personal carbines are still awaiting their first failure. Of the three I mostly shoot, I have a 6920 and two frankenguns made of CMMG parts (with F/A BCGs with black extractor inserts)with Wilson Combat bbls. My current build is going to be mostly BCM parts. I've also considered purchasing an additional 6920 or a DD XV and this leads to me to believe there is no good way to compare the two.
Yes you will. Look at Rob's Chart. The more "X's" next to the companies name means that more effort goes into the product.
You cannot go wrong with BCM and DD.
C4
Yes you will. Look at Rob's Chart. The more "X's" next to the companies name means that more effort goes into the product.
You cannot go wrong with BCM and DD.
C4
I understand that aspect, but even comparing BCM to DD to Colt which is better if we cannot know what parts are going to be rejected; we simply know that they test.
Not trying to be a pain here, just trying to gain a better understanding.
Not trying to be a pain here, just trying to gain a better understanding.
No apologies necessary. This is interesting. I want to know, too.
For example, if testing barrels is SUCH a big deal, then why can't anybody give us a ballpark figure regarding how many barrels fail?! Yes, it DOES make a difference. Are 5% flawed? 50%? 0.5%?
Bimmer
No apologies necessary. This is interesting. I want to know, too.
For example, if testing barrels is SUCH a big deal, then why can't anybody give us a ballpark figure regarding how many barrels fail?! Yes, it DOES make a difference. Are 5% flawed? 50%? 0.5%?
Bimmer
Thanks. These questions have plagued me for some time. This would also provide good insight to the comparable quality (or lack there of) of tier two guns relative to top tier.
No apologies necessary. This is interesting. I want to know, too.
For example, if testing barrels is SUCH a big deal, then why can't anybody give us a ballpark figure regarding how many barrels fail?! Yes, it DOES make a difference. Are 5% flawed? 50%? 0.5%?
Bimmer
Oh, I can take a WAG if you like. I would say that about .3% of the barrels and bolts that are HPT/MP'd fail.
C4
lanceriley
07-20-09, 18:46
why does acceptance rate matter?
if they say that barrels have 50% failure rates... who cares. as long as you're in the 50% good barrel.
Thanks. These questions have plagued me for some time. This would also provide good insight to the comparable quality (or lack there of) of tier two guns relative to top tier.
They don't plague me at all because I'm generally not buying from companies that cut that many corners in the first place. I don't see why people get worked up over this.
Failure2Stop
07-20-09, 19:29
There seems to be an influx of "Tier" speak recently.
The whole "Tier" designation is flawed from the outset, thus the preference here for the comparison chart and observed failures.
Funny thing, the guns that are the most likely to fail in high volume training are the ones with the lowest quality of components and least amount of testing of those components.
Hmmm.
why does acceptance rate matter?
if they say that barrels have 50% failure rates... who cares. as long as you're in the 50% good barrel.
It matters because some companies allow defects. Other companies allow none. Which company would you want to buy a barrel & bolt from???
C4
It matters because some companies allow defects. Other companies allow none. Which company would you want to buy a barrel & bolt from???
C4
By that do you mean companies that properly test their parts like Colt or BCM, or companies that batch test like DPMS and RRA?
I see where tylerw02 is going with this and I would like to find out myself if for nothing more than to satisfy my curiosity, but on the other hand as long as you buy tested parts what does it matter?
lanceriley
07-20-09, 22:51
then don't buy from companies that allow defects.
if that is your statement then the issue is not about defect rate but company reliability.
There seems to be an influx of "Tier" speak recently.
The whole "Tier" designation is flawed from the outset, thus the preference here for the comparison chart and observed failures.
Funny thing, the guns that are the most likely to fail in high volume training are the ones with the lowest quality of components and least amount of testing of those components.
Hmmm.
That is absolutely true. But what are the accept/reject thresholds? They can test all they want without saying where something is too low quality to be accepted. For instance, will BCM and Colt reject exactly the same part or do they accept different levels of variance? Or are we to assume a bolt from Colt=BCM=Daniel Defense?
Once again, because something isn't individually tested, we cannot know for sure if it would or wouldn't pass tests; therefore the rejection rate could be useful to know the likelihood an untested part is flawed. The SWAG guesstimate given above of 0.3% would suggest that three in 1,000 would fail the test. Then one must consider how often the part would fail even if it doesn't meet spec.
That also brings up a good question, of parts failures, how common is it that shot peen, MPI, or HPT would have prevented the failure as opposed to the problem being from something such as the wrong extractor spring/insert, poor staking, .223 chamber as opposed to 5.56, or the wrong buffer? How common is it for barrels to fail? This type of stuff could be very useful to new guys to help them determine WHAT is most important and to know the likelihood of failures of different examples of AR15s. I didn't start this thread to convince people to use inferior parts or to allow apologies for lower tier AR15s; rather to expose what corners being cut will cost the user more reliability and function. What compromises can and cannot be made for the budget-conscious shooter.
then don't buy from companies that allow defects.
if that is your statement then the issue is not about defect rate but company reliability.
All parts are flawed. Quality is a determination of which is less flawed; which has acceptable flaws.
I often wonder what happens to parts that don't pass inspection, but appear cosmetically good.
Are "fail" parts returned to the manufacturer, thrown in the trash, other?
Didn't Pass Military Spec
;)
Most companies if the flaw is rectifiable, will try to re-work the part, as 5min more machine time is more efficient that then X more time to manufacure a new one. However for parts that cannot be fixed (ie. flaws in metal) they are scrapped, and sold as bulk type metal.
FromMyColdDeadHand
07-21-09, 08:30
Just out of curiosity, if parts are batch tested, how likely is it that the test part will differ from the batch? In AR parts, is it that if one part is bad, the whole batch is bad usually?
I guess what I'm aksing is that with AR parts, does one bad apple mean the bunc his bad, and if one apple is good does that mean there are no bad apples?
Been a long time since Stats and running SPC data, but batch testing is more about process control, but indivual testing is more of a guarantee?
We individually test our parts, so I cannot claim to be definitive on batches.
However with CNC machinery, the run down on that machine or fixture part of that machine is likley to be all bad, due to program error, or fixture placement arror. Potentially a tool may wear or break, and some of the run is okay and some not.
If we are talking about Bolts, barrel and carriers, there are so many variables, that batch testing is not a reliable method, unless a large percentage of the run is bad. It really all comes down to stats, and if you batch test, the batch needs to be pretty large.
tyler,
You are putting way too much thinking into this in my opinion. Here it is in Neanderthal terms. One need only look at the percentage of AR's from manufacturer XYZ that routinely take a crap and the ones from manufacturer ZYX that don't.
In other words if someone had told me that Bushmaster was crap based on their name alone I wouldn't have believed them. But, having worked on them and seeing first hand the issues that they have I don't need to be beat over the head. The same goes with DPMS and Oly.
The same holds true for Noveske, LMT, or Colt. If they routinely had issues and were breaking down then we have this information to refute claims of "they are better" than. But, it's not the case.
The common denominator is of course that certain companies actually take the time to build a product correctly and others do not. The ones that don't can't even justify why they cut the corners that they do. Someone made a comment here that Bushmaster sells more AR's than any other manufacturer. That may be true. Now just imagine how many AR's they would sell if they staked the carriers and receiver extensions correctly, used M4 uppers with feedcuts, used the same barrel steel as Colt, BCM, LMT or Daniel Defense, extractor inserts, etc.... At that point people wouldn't have to look hard because the playing field would essentially be even. We also can see that the current price difference is only about 100.00-350.00 in most cases.
To me that means that company XYZ is more concerned about their profit and not about the product they produce. Now let's look at the parts. Why would any of those aforementioned companies be overly concerned about parts failure or the quality of their parts when they don't even care about the platform that they build? Not to mention all too often (we see it here all the time) someone is asking how to fix their new out of the box DPMSBUSHOLYMODEL1 Ninja gun rather than contacting the manufacturer and making them fix it. If everyone did that they would eventually have to change their ways or they would fall by the wayside.
DISCLAIMER- If you (general descriptive term) are butt hurt about what I have said too bad. It is what it is.
Thanks. These questions have plagued me for some time. This would also provide good insight to the comparable quality (or lack there of) of tier two guns relative to top tier.
By that do you mean companies that properly test their parts like Colt or BCM, or companies that batch test like DPMS and RRA?
I see where tylerw02 is going with this and I would like to find out myself if for nothing more than to satisfy my curiosity, but on the other hand as long as you buy tested parts what does it matter?
Both. BCM (for instance) has a ZERO flaw acceptance policy. So if the LAB finds ANY issues with the bolt or barrel, it is rejected.
C4
I often wonder what happens to parts that don't pass inspection, but appear cosmetically good.
Are "fail" parts returned to the manufacturer, thrown in the trash, other?
Ah, this is a good question and one I have been waiting for.
So let's pretend that you have a .Mil contract. The Govt has given you the TDP to follow. A Govt inspector comes in and takes measurements on say a bolt. It does not meet the TDP spec, but is 100% serviceable and may actually be a better fit than the bolt following the TDP (which might be on the low end of the spec).
The manufacturer will then use that part in their LE or Civy AR's (or sell it). Since this part did not meet the .Gov standard, should it be classified as crap? No, not in the least, it just means that it does not meet the blueprint drawing that the Govt is using for that part.
Now this does not mean that ALL parts that fail a .Gov inspection are still quality. In the case of the bolt though, it does mean that it is still useable.
C4
tyler,
You are putting way too much thinking into this in my opinion. Here it is in Neanderthal terms. One need only look at the percentage of AR's from manufacturer XYZ that routinely take a crap and the ones from manufacturer ZYX that don't.
In other words if someone had told me that Bushmaster was crap based on their name alone I wouldn't have believed them. But, having worked on them and seeing first hand the issues that they have I don't need to be beat over the head. The same goes with DPMS and Oly.
The same holds true for Noveske, LMT, or Colt. If they routinely had issues and were breaking down then we have this information to refute claims of "they are better" than. But, it's not the case.
The common denominator is of course that certain companies actually take the time to build a product correctly and others do not. The ones that don't can't even justify why they cut the corners that they do. Someone made a comment here that Bushmaster sells more AR's than any other manufacturer. That may be true. Now just imagine how many AR's they would sell if they staked the carriers and receiver extensions correctly, used M4 uppers with feedcuts, used the same barrel steel as Colt, BCM, LMT or Daniel Defense, extractor inserts, etc.... At that point people wouldn't have to look hard because the playing field would essentially be even. We also can see that the current price difference is only about 100.00-350.00 in most cases.
To me that means that company XYZ is more concerned about their profit and not about the product they produce. Now let's look at the parts. Why would any of those aforementioned companies be overly concerned about parts failure or the quality of their parts when they don't even care about the platform that they build? Not to mention all too often (we see it here all the time) someone is asking how to fix their new out of the box DPMSBUSHOLYMODEL1 Ninja gun rather than contacting the manufacturer and making them fix it. If everyone did that they would eventually have to change their ways or they would fall by the wayside.
DISCLAIMER- If you (general descriptive term) are butt hurt about what I have said too bad. It is what it is.
No, nobody is "butt-hurt" as you put it. I'm looking for answers and it appears not getting too many. Let's break down some of the things you spoke about
"Now just imagine how many AR's they would sell if they staked the carriers and receiver extensions correctly"
I agree, they should do this. This is a quick and simple job and should NOT be overlooked. Meanwhile, it shouldn't cost hundreds more for X company to do it, either. Staking an extension can be done in about ten seconds. Staking a carrier can be done in just a few minutes. There is no reason why the junk manufacturers shouldn't do it. These are very critical for the operation; probably much more so than whether or not a bolt is MPI tested---which is precisely why I ask how common it is for a bolt to be rejected. If, for instance 1 in 1,000 units would fail the test, odds are a non-tested part doesn't mean the part wouldn't pass and you have a junk part.
"used the same barrel steel as Colt, BCM, LMT or Daniel Defense"
Once again, how common is it that 4140 fails and 4150 or Mil-B-11595E+ would not? Most private citizens are using semi-automatic carbines. Per gotm4's description on this sight there is a "slight" difference. Throats aren't going to last much longer with one or the other it seems. It appears that 4140 is "adequate" for semi-auto use and Mil-B-11595E+ is icing on the cake.
"extractor inserts, etc."
It is my understanding blue extractor inserts are for rifles and black are for carbines. Is that assumption correct? Having the wrong extractor isn't a major item to change.
"Not to mention all too often (we see it here all the time) someone is asking how to fix their new out of the box DPMSBUSHOLYMODEL1 Ninja gun rather than contacting the manufacturer and making them fix it. If everyone did that they would eventually have to change their ways or they would fall by the wayside."
I don't think we'll ever see the day because most people are tight-wads and care more about putting a China red-dot sight and UTG quad-rail and making the firearm look cool than they are about the details. Most people aren't going to get educated as to what will ensure their firearm continues to run properly.
I didn't start this thread to compare things like inserts and staking, rather the likelihood a non-tested part would fail. Any bubba with his StaPMSBushStar1 could have his buddy stake his carrier key with his MOACKS, stake the receiver extension, install the proper $4 insert/spring, install the proper $12 buffer, and get his .223 chamber throated-out for 5.56 ammunition and make the over-all carbine much more reliable for a very minimal cost. If those alterations are made, whether or not the manufacturer does it or not--that is irrelevant, what are the odds that the barrel, bold-carrier, or bolt would fail? Just because a part isn't tested, doesn't mean it would fail so rates are pretty important. What are the rejection rates of parts from the top tier?
This is not a flame or a "just as good as" thread, rather wanting to put some real empirical data with the concept of shot-peen, MPI, and HPT. It appears that isn't going to happen.
Some time ago, a rep from a lower end manufacturer inadvertently stated that the parts they rejected for their complete guns went to their parts kits. The justification being that parts guns were more likely to be inexpertly assembled and have other problematic components, giving the manufacturer some deniability if problems occurred. My kingdom for a voice recorder on that one.
There are signs of the same practice in another lower-end manufacturer. Upon information and belief they do maintain a standard, albeit a pretty low one.
In the Colt AC, Ken Elmore often noted the number of requests Colt would receive from other manufacturers for Colt's rejected parts. There were no surprises.
I really love these types of threads. Right next to; what kind of AR should I buy and why? :rolleyes:
I really love these types of threads. Right next to; what kind of AR should I buy and why? :rolleyes:
This isn't a "what gun should I buy" rather a "to what degree should I avoid"...I'm looking for concrete data.
:rolleyes:
Feel free to contribute if you can answer the questions. If not, why flame?
CaptainDooley
07-21-09, 11:07
I guess I just don't understand the fuss... Instead of playing a crap shoot with lower teir parts and trying to justify it with statistics, why not just buy quality, tested parts? Personally my life and the lives of my family aren't worth the gamble.
I guess I just don't understand the fuss... Instead of playing a crap shoot with lower teir parts and trying to justify it with statistics, why not just buy quality, tested parts? Personally my life and the lives of my family aren't worth the gamble.
Nobody is trying to justify anything. I'm trying to get some empirical evidence rather than the vague concept of "tested" vs "batch-tested" vs not tested.
I never said it was a "what gun should I buy" discussion. I said it's rather silly JUST LIKE those threads. That old saying "the proof is in the pudding" applies here.
This isn't a "what gun should I buy" rather a "to what degree should I avoid"...I'm looking for concrete data.
:rolleyes:
Feel free to contribute if you can answer the questions. If not, why flame?
Jay Cunningham
07-21-09, 11:36
I believe it would advantageous for M4C members to know the actual rejection rate of tested parts to be able to suggest the likelihood that non-tested parts don't meet spec and are prone to failure.
I suppose I won't find good organized empirical evidence to suggest brand X over brand Y.
Not trying to be a pain here, just trying to gain a better understanding.
I didn't start this thread to convince people to use inferior parts or to allow apologies for lower tier AR15s; rather to expose what corners being cut will cost the user more reliability and function.
No, nobody is "butt-hurt" as you put it. I'm looking for answers and it appears not getting too many.
This is not a flame or a "just as good as" thread, rather wanting to put some real empirical data with the concept of shot-peen, MPI, and HPT. It appears that isn't going to happen.
This isn't a "what gun should I buy" rather a "to what degree should I avoid"...I'm looking for concrete data.
Nobody is trying to justify anything. I'm trying to get some empirical evidence rather than the vague concept of "tested" vs "batch-tested" vs not tested.
tylerw02,
You've gotten some good information from some industry professionals. As we can all see from the trend in your above statements, none of this information is apparently good enough for you. I suggest contacting all of the major AR-15 manufacturers and asking them for the numbers (no doubt guarded information) which you seek.
Sticking with your same argument ("not good enough, not good enough, not good enough... etc.") is going to be viewed as trolling.
I never said it was a "what gun should I buy" discussion. I said it's rather silly JUST LIKE those threads. That old saying "the proof is in the pudding" applies here.
That's what I'm looking for---some proof. Most AR failures I've personally seen (and even read about) can be attributed mostly to magazines, lubrication, inferior/wrong extractors, the wrong buffer, and poor staking. I'd like to find a source with some hard data on parts. It appears that isn't going to happen.
That's what I'm looking for---some proof. Most AR failures I've personally seen (and even read about) can be attributed mostly to magazines, lubrication, inferior/wrong extractors, the wrong buffer, and poor staking. I'd like to find a source with some hard data on parts. It appears that isn't going to happen.
What do you expect us to say? No one in this industry is going to give you their scrap rate.
Secondly if that all you see fail, you need to do more shooting ;)
Powered by vBulletin® Version 4.2.5 Copyright © 2024 vBulletin Solutions, Inc. All rights reserved.