Three vids posted on YT, this is the one for lot testing.

 

YT videos of Lapua Rimfire Production from 8 years ago

Curious does anyone know of any newer videos available? Only so much you can glean from watching this. Three things stand out:

2. Test fire at 1:17 Looks like Anschutz barreled action. Very likely easy to change barrels. This appears to be a barrel

Yes, to my old eyes it looks Anschutz and an older rifle, the bolt has a flat cap (earlier) compared to newer ones which are rounded/domed shaped.

Then again, I have little or no experience with new modern Anschutz rifles as mine are all old models.

Vids like this from Lapua or Eley Tenex will not give away any trade secrets, alas those secrets are not for the end users.

 

It doesn't appear to be an Anschutz action, old or otherwise.

 

It doesn't appear to be an Anschutz action, old or otherwise.

I stand corrected, it's these old eyes of mine, ..............getting old sucks.

 

Walther action..kk200 or kk300.

 

That's what I was thinking too. Not that I know much about them. I just didn't think it looked like an Anschutz action and the Walther is a very likely candidate

 

Here's another SK/Lapua video. A few more snippets of factory production and one operation which is of interest, the primed (inert propellant, not spun in, compressed) propellant filled cases getting dosed with "activation" water/acrylic binder, "milky color fluid". Eley Patented this process and possibly SK/Lapua licensed its use years ago although Eley's patent has long since expired.

The SK/Lapua factory in Schonebeck has been around for quite some time......

In 1829, the merchants Louis Sellier and Nicolaus Bellot founded a company for making percussion primers. In the 150 years to follow, the company developed into a significant producer of ammunition and explosives. In 1992, Lapua Oy, today the Finnish enterprise Nammo Lapua Oy, bought the ammunition production in Schönebeck in Germany. Under the guidance of the new owner, the successful production of subcaliber, hunting and shot ammunition of the traditional SK brand could be further extended and developed.

 

Where was the part about production in the OPs video? It started out with packaged boxes of ammo, no production of it at all. Showed a little QC & test firing but no production.

 

Why does the original video have the continuous thump thump thump bass percussion in background? A lot of radio ads now have this. Am I the only person who finds this annoying enough to turn off the ad or video?

 

I also find it annoying, I just turn down or mute.

 

@NIB

Thank you for the video addition! Here at :24 you can see the Wather KK 300 is the test rifle used for this Center-X lot.

 

SK factory snips. Could be some duplicate equipment seen in the other video. More of that "drumming" music, mute and add your favorite music to avoid.

 

Nammo Schonebeck short video. Schönebeck - Nammo | Nammo

 

Then why did the 5 lots of Xact i tested shoot no better than the Center X??

 

There are several possible explanations. An obvious one is that how the grading is done is not based exclusively on shooting results, on how the ammo performs. Perhaps it's only partly based on how the ammo shoots, perhaps not at all. In other words, the lots that come off the production run that are supposed to be X-Act get labelled as X-Act.

Below are some questions for speculation. The ammo makers don't reveal how the grading of ammo is achieved.

A single production run may produce lots of X-Act, Midas, and Center X. Is it likely that when a production run is made that produces tens of thousands of rounds (perhaps more?), which are divided into lots based on various criteria, that the ammo makers have absolutely no idea of what the resulting products will be? Is it possible that through decades of experience they have a reasonable idea of what the end product will be like -- even without testing by shooting?

All the components that go into the product are themselves graded -- the brass casings, the powder, the bullets. Is it likely that they are put into the loading machines willy nilly? Is it more likely they loaded so that the best components come together in the loading process to produce lots that have the top-graded casings, powder, and bullets? These could be labeled X-Act. The next best graded components could become Midas and the rest Center X.

If the loading process has known variation in output quality, which after decades of experience would be understood, could this too could be used to advantage? For example, if the beginning, middle, or end of the production run has certain proclivities to produce a better product, could that be used to advantage?

 

I was just at the Lapua test center in Mesa and ended going with Midas. stared with 12 lots 8 lot didn't work in my 2500X Muller barrel. 4 lots shot good. In the 2nd round of testing 2 lot stood out with 1 being fractionally better. The lot I picked shoot the best in my 2500X Shilen barreled rife. It all comes down to testing good ammo for good results.

 

You would have to be a fool to grade ammo based on any parameters other than how it shoots.........factory, test center, or shooter.

 

It still begs the question of how very poorly shooting lots with the name X-Act or Midas get to market, doesn't it? No barrel can fix inconsistent ammo.

A grading process based on factory shooting results should avoid the problem of lots of top graded ammo that don't shoot well in multiple barrels. Nevertheless there are poor lots of all grades of match ammo.

It's not difficult to understand that lousy lots get produced. The question remains how they get to shooters.

 

Lot grading is based on shooting samples of each batch
from multiple fixtured barrels at in indoor facility.
Best results are labeled accordingly.
Specifications are proprietary.
Allowable spread, number of strays, number of cartridges tested
are all hidden from the consumers. Batch labeling
is established by statistical sampling, with confidence level,
calculated percentage error allowed, determines sample size.
In a business, profits are job one. If quality testing
affects the acceptable cost per cartridge, corners are cut.
Small sample sizes lead to incorrect conclusions.
That is how lots end up shooting above or below their price point.
Failure of the testing process to use a large enough sample
to provide an accurate assessment of the entire batch.

Close enough for government work.

 

"Small sample sizes lead to incorrect conclusions."

If you are using "statistical" samples, valid samples is what I assume this to mean, then you can't have a "small" sample issue. If the error is due to sampling you essentially have an inappropriate sampling methodology. Assuming a 5% error allowance you will have 5 "bad" samples per 100. Anything more and you are not using a valid sampling methodology for your population.

 

Testing small samples can indeed lead to incorrect conclusions.

No one complains when the ammo is as it should be and this is what happens most of the time. In these circumstances there seems little to be gained by testing large samples to verify what is anticipated. When testing is done smaller samples are adequate.

When the ammo lot is better than expected, shooters rejoice, don't complain and are pleased to get CX that shoots better than some M+ (or even some X-Act.) In such circumstances, it's unlikely that grading was by testing only.

It's when the ammo turns out to be not as good as expected that testing small samples can fail the buyer.

The factory testing barrels don't have some magic that makes them shoot poor lots of X-Act or Midas shoot much better than they do in other barrels.

 

So does the factory grading process have a failure rate, where despite their procedures they somehow get it wrong? If so how does that rate compare to the failure rate of the end user?

 

I've asked Lapua, Eley and RWS about their grading process.
The reply from each was "proprietary information."

Common to each company's description of the process
was fixtured barrels/receivers, indoors, samples from each run fired,
results recorded, batches labeled according to predetermined standards

No information regarding sample size per number of cartridges per batch.
Nothing about the brand's standards for labeling.
No information regarding level of confidence of testing.
No information regarding acceptable defect percentages.

I don't see that as trade secrets, but deliberate concealment
to avoid consumer reactions that could affect profits.
If we were to learn that level of confidence was only 80%,
plus or minus 5%, then on a bad day, the results
would only be repeatable 75 out of 100 times.
If the acceptable defect rate is 1% per thousand
you could be sold 50 out of spec cartridges per case.
That would explain substandard boxes showing up in consumers hands.
What concerns me regarding the process,
is when entire batches are voluntarily recalled,
after supposedly having been tested by the factory.
Look at the original Eley Long Range fiasco.
Substandard results from what was supposed to be
the new, best option, for extended range accuracy.
Does that instill any confidence in factory lot testing?

Makes you wonder how low the factory standards are,
how small the sample size is per 25,000 cartridges,
how large the acceptable defect rate is,
and, as Penage Guy likes to remind me,
do they actually test every batch?

I found a factory batch testing target on line.
200 cartridges were actually tested, of a run
that had over 28,000? cartridges in it.
Is that an acceptable sample size?

Side note, Eley would publish the factory testing targets
on the Eley website, to allow consumers
to decide which lot to purchase.
That was discontinued after a short trial.
I wonder why? Did it affect sales?

Based on the above target and an online sample calculator...

So for a 28,000 cartridge run, 85% level of confidence,
+/- 5% acceptable error, 206 cartridges need testing.

I have no experience with quality control used in manufacturing.
What I know about statistics and their applications
goes back to courses taken in the late 70's.
Barely useful in current manufacturing and production.
Those that do have experience, is 85% level of confidence acceptable?
85 out of 100 attempts will obtain the same results?
15 out of 100 tests from the same batch of product
will produce substantially different results?
That's a fairly hefty difference, and would explain why
there are so many lesser quality cartridges showing up
when purchasing what are supposed to be the best available.
It also would explain finding cartridges
producing results well above their price point.

 

I found a factory batch testing target on line.
200 cartridges were actually tested, of a run
that had over 28,000? cartridges in it.
Is that an acceptable sample size?

One can't tell if it is acceptable from this information alone. It depends on how "error" is dispersed in the population. It depends on how extreme the "error" is. It depends on the number of elements that affect the presentation of the error. There are other factors as well. Error in the real world is usually multifactored. Not due to one thing but due to the interaction of many, often very subtle, things.

Dispersion: Most traditional parametric statistics assume fully randomized error. For example, if error occurs in clumps or patches then you need a different sampling approach.

Error Intensity: If there is a a big difference when there is an error it requires less sensitivity to find it.

Error interactions with testing equipment: This is usually the result of not understanding the "causes" of the error. As such the equipment may be sensitive to one form of error over another.

Multifactorial error: Rarely is there a simple single error element. Even if the various causes of error are fully randomized, the fluctuations of multiple random elements impact detection and diplay. Random stacked tolerances are a form of this. Random variations in case thickness, primer amounts, power amounts, bullet weights, etc. create situations where one element being "low" is offset by another being "high."

We can go on and on here.

Sample size is complex. Where big impacts of error are easily detected a sample of 1 is sufficient. Death, for example, has many "causes" we can detect with a single subject as the sample.

 

Thanks jaia.....Still 10 times the rounds tested vs the two 10 shot groups you get at a test center when you buy your case. Also who here at RFC ever tests their ammo out of 4 rifles with five 10 shot groups each with computer analysis. Little room for complaints from most shooters, the factories work harder at it than we do.

 

Alot of speculation here it seems..

Their not going to tell the public how they grade their ammo or what the quality control is.. or any other aspect of their ammo.

We're fortunate that they give us a average velocity from their test barrels..

Its up to the consumer to test their ammo.. to grade it in our rifles and determine the precision we seek that will satisfy our needs..

 

Alot of speculation here it seems..

Their not going to tell the public how they grade their ammo or what the quality control is.. or any other aspect of their ammo.

We're fortunate that they give us a average velocity from their test barrels..

Its up to the consumer to test their ammo.. to grade it in our rifles and determine the precision we seek that will satisfy our needs..

I don't know about that. If a company tested more, and could provide each box with specific info, and if they could show it exceeded what other companies do, then it could become a selling point. A niche market, but it would be a very big selling point.

OTOH, never have I found anything so difficult to get across to lots of people than simple statistics. The math isn't really bad, the concepts are not difficult though sometimes unintuitive, but somehow it is hard to get across. So maybe a company does what I describe above doing extra testing, providing the info on that testing and consumers who don't grok statistics fire a couple 5 shot groups and declare the more thoroughly tested ammo not very good.

I'm not incredibly impressed with the testing at test centers. Great conditions to test in, but I think the number of shots fired is not enough to be finely discriminating of close results. It is enough to rule out bad choices, and to get good results. The underlying issue may simply be the variability inherent in rimfire. Maybe they have tested enough to know that results of their best stuff is too variable to gain anything with more testing. Something they are almost surely not going to admit.

 

Here are some pictures from the Lapua Test center in Mesa on July 22. I ordered a case of Midas+. The case I orded shot very well in both my 2500X's one has a Ratch barrel the other Muller 8 groove. The Muller is in the picture.

 

What level of confidence do you expect from your testing?

That is the deciding factor.

50%? We have that now.
I call it the assembly line lottery.
I purchase random bricks when available on line.

85%? That means 85 times out of 100
testing will produce the same results.
15 times out of 100 the results won't be the same.
Is that acceptable?

How about 95% level of confidence?
Do you know how many cartridges would have to be tested?

So to purchase a case of match ammo
at a factory test center, how many cartridges
would need to be tested to trust your decision?

On the humorous side of the process,
think about the laughter generated from folks
who work quality control and statistics every day,
hearing about buying 3 cases of ammo,
based on 40 shots actually fired.

So what level of confidence and batch size
needs a 40 shot test sample?

Now I know why they're laughing.

 

Jaia, you just take all the fun out of going to the test centers.😁

 

John, until this morning I never thought what sample size of 40,
would apply to. I had to trial and error the calculator
in order to find a combination that worked.

I've been reading about quality control and statistical sampling,
after a few messages from folks here at RFC
directing me to on line publications and calculators.
It's been educational and improved my understanding
of why there can be so much variation in quality
even after batch grading at the factory.

 

Very true. At the test center in July I had one lot with the first 10 rounds group under 10 at 50 meters, not yards. I thought "the holy grail" lot. Next 10, same box, was 15.5.

 

A related question about match ammo production is how often what comes off the production line is what the ammo maker expects?

Since the ammo makers don't reveal how ammo is graded, the following is speculation. It offers an explanation of why testing can be minimal. It offers an explanation of why sometimes top tier grades like X-Act can be poor and why sometimes third tier can be better than expected.

Consider that the ammo makers grade all components that go into the ammo, the brass casings, the propellant, the priming compound, the bullets. These components are not all identical, some being better than others.

A purpose of this grading may be quite practical: to get the best components together in the same rounds to increase the chances of producing better products -- i.e. better rounds.

The grading of these components may well be so that the best of them can be loaded together, the top-graded brass casings with the top-graded bullets and, however they are evaluated, certain powder and priming compounds. The next "grades" of components can be loaded together and so on.

Perhaps it's not unreasonable to expect that, after decades of match ammo making experience, it is likely that the ammo maker has learned what to anticipate about the quality of ammo batches being produced during a production run.

Over the decades, .22LR match ammo makers don't produce tens of thousands of products (rounds) each day (dozens of cases) without learning something about the likely quality of the end products. They must get it right most of the time. Most of the time what they expect to be very good quality match ammo is indeed just that. Everything else falls into place with second and third tier products too.

Sometimes in the course of production, with the vagaries of human involvement and dozens or hundreds of moving parts, the products are not as expected. That's when some lots of X-Act are very poor by any standards, correctable by no barrels. That's when some lots of second or third tier product outperform the top.

But most of the time, the output of the production line is as expected. Vigilance and care is always needed. Perhaps a little testing helps too.