Was doing some searches on annealing and ran across this, was wondering if anyone here has tried it and thoughts... seems labor intensive for the amount of annealing I may be looking at, but found it interesting none the less...

http://www.65creedmoor.com/index.php?topic=6019.0

Exactly this.

Molten salt is a bit more scary than molten lead, at least for me.

I’ve managed to anneal a bunch of cases with the method dryflash3 describes, with a $10 (and cheap!) rechargeable screwdriver a $15 propane torch, a $3 deep well 10mm socket and $24 worth of Tempilaq from Midway.  I already had the “1/4” hex drive to 1/4” socket drive” adapter.

I learned that standard propane torch type tanks have the same fitting as propane camp stove tanks, so the excess camp tanks I have squirreled away will keep me annealing for a long time.  And since it only takes a tiny bit of Tempilaq for each case, I won’t need another bottle for a long time.

I came across that same thread a couple months ago and I decided to give the salt bath annealing a try.  Here is my setup:


I put this together for like $50.  I bought some potassium nitrate and sodium nitrate and mixed my own annealing salt.

So far I like it and think it works very well.  I don't know how many cases you are looking to do, but it actually goes pretty fast.  I can do two cases every 10 seconds, so 100 cases takes just over 8 minutes.  I use my phone to play a 5 second interval timer video from youtube to set the pace.  On the first beep I drop a case into each hole, on the second beep I pull them out and drop them into a bucket of water (that is well away from casting pot).  Lather, rinse, repeat.

As far as safety goes, 900 degree molten salt is NO JOKE, which should go without saying.  I feel pretty safe doing it though with only two small exposed holes to drop the cases through.  I actually drop the cases in and pull them out with my bare hands.  I'm less likely to slip and I like to make sure the case head stay cool (they don't even get warm).  While this system is certainly more labor intensive than a automatic annealing machine, its a LOT less expensive and if anything more consistent.

@thejessman

Thanks - I think this is awesome!!

I also like the way you set up the molten salt tank with just 2 holes in the cover; that looks like it would seriously limit the anticipated danger associated with molten salt.

RE salt bath heat treating: this technology is also used in knife-making; the temperature of the "liquid" salt is extremely consistent throughout the bath and is easier to maintain than a kiln (where a gas - ie atmospheric air - is the medium).

I made a post a while back in the AK section of the forum (in build it yourself) about applying this tech to full-receiver heat treating; but it didn't get much traction.

Q. For you: is that a lead-casting pot you used?

Yes, it is a Lee lead casting pot.  I bought it specifically for this purpose, they are around $35.

I can't take credit for the design of the cover/case-holder.  I pretty much directly copied the design from the 65creedmoor site.  When I first melted down the salt I sat there staring at the open pot of molten salt and thought to myself "Yeah, this is scary".  But after I put the cover on and started using it I felt better about it.  The cover doesn't remove all the danger, but it greatly reduces it.

Only $50, okay, but does it do anything?

What I mean is, do annealed case necks ACTUALLY improve accuracy?  I have not seen any effect in my testing.

Can some post data that shows a statistically significant improvement (reduce group size)?.

In my Dillon XL650 vs. Lee Classic Cast thread, I will be looking at the effects of annealing on runout and muzzle velocity dispersion

Reuse of cases seems to be the big benefit.  I am hoping to do 308 to keep cases running in an M1A longer than a couple loads

A few years ago, we had a thread on here about fatigue life and the effect of annealing.

IIRC, there was no significant increase in case life.  There should have been, one hopes there'd be, but IIRC the data did not support that conclusion.  If someone can resurrect that thread, that would help.

If someone can post improved accuracy data, that would also help.

I thought about salt bath annealing before making a more traditional torch setup. I was hand annealing with a torch and dropping into water afterwards (and dipping my hot fingers in there too occasionally.)

I was corresponding with someone from another forum that has a friend that designs/implements annealing equipment for government contract ammunition facilities. We were discussing inductive annealing and getting input from his "expert" friend. Inductive annealing is what these commercial operations use.

During this discussion, his expert friend said we should not be putting the case into water after annealing. The case needs to come to room temp naturally to obtain optimum results. I took this piece input as gospel because of his experience.

For this reason I veered away from the salt bath method, because a water dunk post annealing was mentioned in the process. (I suppose you could decap, salt bath anneal, let air cool and then wet tumble...but didn't cross my mind at the time.) The salt also needs to be added after a number of cases are annealed to keep the bath height at the required level. As the level drops, the heat transfer to the case would be altered from the first to the last before refilling the bath so a method for maintaining count of the cases through the system would need to be used since you can't see the level without removing the lid. These were just my OCD thoughts at the time...

The inductive method has a similar workflow as the salt method, but processing only one case at a time instead of 2. Inductive was more complicated to build, more expensive and a lot more difficult to automate...so I built an automated torch machine instead.

I'm sure you'll get good results from the salt method, I just took different direction for the reasons above. Good luck!

I haven't done any accuracy testing, it would be interesting to see.  I anneal to hopefully increase case life, but also to improve neck tension consistency.   When seating bullets I can feel the difference in neck tension between annealed and non-annealed cases,  I can't measure that difference in feel though, so it could all be in my head.

An induction annealer would probably be the best, but its about the "bang for your buck" and I can't justify the significantly increased cost and complexity of a induction annealing machine.  For others it might make more sense.

You are correct about the level of salt dropping and you run cases though the annealer.  Its not very much though.  Like 1/16" for every hundred cases.  I just add a little spoonful every 100 cases to keep it where it want it to be.  Its actually quite easy to check the level of the salt.  I just dip an allen wrench in, the salt sticks to it when it comes out so its easy to see the level.

Thanks for taking the time to post that and continuing to answer questions... I ordered the pot and temp probe and going to emulate your setup...

Annealing isn’t about “accuracy” as such.  It’s about doing something to reduce or control the work hardening that seating the bullet then firing it does to the case neck.  Hardened necks lose tension, and not in a consistent way, so that DOES affect accuracy.

Annealing also helps prevent case mouth splits, and thus can greatly increase the life span of a case.

Prove it!  Show me the data, that's all I'm asking.  See my posts, above.

Induction is by far easier, 'cleaner', more consistent, and safer. A homemade setup isn't terribly expensive for the added benefits. ($150-$200)

I did a series of accuracy testing with various charge weights both annealed and not annealed. I think it was 10 rounds per charge weight, with a half grain spread, each charge weight was .1gr, for a total of 50 annealed 50 non annealed. This was 75gr match prepped ammo in my match rifle.

Accuracy difference on paper at 200yrds was almost non existent. Sometimes the annealed was tighter, sometimes non annealed was smaller.

All rounds were also chronographed with a lab radar. This is where I did see an annealing advantage. Non annealed averaged around 30fps ES. Annealed was around 20-25fps ES. I did have one non annealed group with 15fps ES, so there was no real 100% clear winner.

However, I also tested 80gr at 300 yds. I was able to fire 5 round 1.5"(.5MOA) groups with annealed brass. The best I could do with non annealed was a little over 2".

Brass life has been about the same with both methods, but I also don't push brass to the max. Split necks are about the same. I usually load 6-8 Times and see around 30 split necks per 1000pcs by the end of the lots life. If you really want no split necks get an Ackley chamber, but that's a whole 'nother story.

The lower average ES I saw with annealing is enough for me to do it for match brass, but not practice brass. It should be translating to more accurate ammo, which might be worth a point or two in a match. It defiantly doesn't hurt to anneal, but I think some of the benefits are a bit exaggerated by the people selling annealing units, but benefits do exist.

Do you have the specific SD numbers from your annealed vs. non annealed test? ES is irrelevant, just need SD numbers and number of rounds for each sample.

It has the advantage of being much less dependent upon exposure time.  Using a propane torch, the window for how long a case can be exposed requires constant attention (or an expensive machine).  Five seconds in the flame may anneal the brass perfectly, but six seconds may ruin the brass.  Substances like Templilaq can help you judge when the exposure times have been long enough, but it must be applied to each case and allowed to sit a specific amount of time.  Feeding cases one after another into the holder is not going to be much more labor intensive than painting Tempilaq onto a case and then getting out a torch to heat it up.

In reference to preventing case splits, I said “can” rather than “will”.  And logically, if you do everything you can to make your cases as uniform as possible, that should remove variables that can be detrimental to accuracy.  I haven’t done enough annealing of cases that get shot a lot to have personal data.

I HAVE annealed my cases in the process of converting .223 to 300 Blackout.  My observations are that annealing those cases before forming them results in less effort to form, and they appear to have less issues with neck thickness than the few I formed without annealing.

I believe that these cases are easier to form, and display more uniform neck thickness because the annealing process relieves stresses from the original case’s production and firing, allowing the brass to respond better to the forming and inside neck sizing by the expander.  Whether this is the case or not, it seemed much easier to form the annealed cases and the hundreds I’ve done this way have appeared to be very consistent.

Work hardening DOES happen from firing and sizing.  And hardened case necks don’t have the same elastic properties as non-hardened necks.  This affects neck tension, and consistent tension provides more consistent combustion conditions within the case.  Cases with random amounts of neck tension are typically not expected to promote accurate handloads.  If a collection of cases has varying amounts of neck tension, they probably will not perform very consistently as loaded rounds.

Whether or not it leads to measurably improved case life or accuracy of handloads, treating and processing cases in a consistent manner reduces variables that can impair accuracy.

So I don’t have any solid data to back up my earlier assertions.  I was really just repeating the information I’ve read - from experts and seasoned veterans of handloading, people whose credentials exceed ‘writing for a gun magazine”.  But my real experience with annealing cases as I was converting them indicates that annealing does something useful, and the research I’ve done on the metallurgy of brass indicates that annealing is useful, if not necessary for every case all the time.

We had a thread on that a while back, as I recall.

https://www.ar15.com/forums/armory/-/42-422324/?

Damned photobucket - I know I've got the graphs squirreled away somewhere, let me see if I can dig them up.

ETA: Here toy go

Precisely!  That is why I am challenging people to prove it.

This parroting what others have said is known (disparagingly) as "an old wives tale" or possibly even "the cargo cult".

I understand all the mechanical engineering and materials science behind annealing.  It's like I want to believe, however,...

My recollection of the last report I saw was it made no significant difference in case life.  This was surprising but was the result.  IIRC, on average it made some small difference.  IIRC it was one additional loading out of 10 nominal reloading cycles.  So, all that cost and all that additional effort to extend case life by one cycle.  This was true for several different case brands tested.  All the above is subject to the vagaries of fading memory.  That's why it would be good if someone could search the archives.  I do not recall any accuracy data in that report.

I'm not sure I've seen any accuracy data with/without annealing.  Honestly, in terms of practical accuracy, I'm not expecting any improvement at all.  If there were to be a demonstrable, significant reduction in group sizes that might be a reason to anneal.

RocketmanOU,

That is the data set I was trying to recall.  I once sat there and re-entered all that data so I could extract the practical info from it.

What do the various symbols mean in that final plot?  IIRC, it shows the mean value (square), the 3sigma limits (bars), min and max values (X),...

Note - there is no significant different between annealed and not annealed until well after 13 loadings at which point half of ALL original cases cases are already lost.

What you are getting with annealing is the area between the non-annealed and the annealed curves.  That number of "shots fired" is all you get.  As a percentage of total shots, it is a small fraction of total shots fired (~20% more total shots of a rapidly diminishing batch of cases).

Also note, during the first 11 firing, more annealed cases were lost than non-annealed.

My data set:

I have a 100 lot of 223 brass that has been reloaded 15 times so far, has been annealed each time, and haven't lost any (brass catcher), or culled any during inspection.  
They have been loaded with the same powder (H322) and bullet (50g Zmax) each time for prairie dogs, they are loaded for accuracy in my AR, they are not loaded to max.
Annealing helps me maintain consistent neck tension throughout brass life, which leads to consistent accuracy, and consistent point of impact throughout brass life.  (which means same poi with brass that has been loaded fewer times ....this is important to me = prairie dogs.)
Obviously I am getting good brass life annealing, and accuracy is great.

I do not oversize, they are sized for one AR, and have been shot in it exclusively, including the first time (Lake City Brass), they were loaded when I bought/shot them the first time.

Additionally, I have 500 lots of other manufacturers headstamps sorted/segregated/reloaded (and other 1000 lots of lake city), that are holding up very much the same as the above that have been annealed each time since I got my bench source.   They don't have 15 reload cycles like the ones above (that's a lot of shooting), but I have seen nothing so far that would leave me to believe they won't make it there (minus a few lost along the way I am sure)

Additionally, I also believe in my stuff, annealing helps me freshen up neck tension/produce once fired range brass (head stamp sorted to Lake City), that shoots the same load virtually the same, when I choose to (which isn't often these days, I have enough brass).

For those who want to say annealing does no good, or isn't worth the time, does nothing to help create consistent neck tension, does nothing to aid brass life, does nothing to prevent work hardening, no problem, do (or in Troll's case, don't do) what works for you, in your stuff, for your purpose....no problem.

There is no one data set that will prove annealing benefits (or lack thereof) for everyone else, because in most instances the tests are flawed (if we are talking true scientific tests)....but the biggest flaw is the tests are done with someone else's equipment, someone else's reloading techniques, someone else's load, and shot in someone else's equipment.  The data set that matters (to me anyway), is what works for me, in my stuff, for my purpose.   Meaning, if I was reloading to berm blast and/or shooting where I would loose brass anyway, I would be less interested in annealing (although it's so darn easy with my annealer, I probably would anyway lol).  I am experienced enough to know there are few absolutes in this game, ymmv

Troll, you asked for data.   The above is mine.   Now, where is your data that annealing doesn't work in my stuff?

Back to the OP:

Thank you for introducing the salt bath thing.   Admittedly it is unlikely I will change or look at alternatives unless and until my method quits working for me for my purposes, but I haven't paid any attention to salt baths (beyond some funky smelling bath stuff my wife has around) until you mentioned it.   Thanks for that.

I will add this: annealing is not some voodoo that magically makes your random loads laser-accurate.  It is a technique that can restore neck tension to cases that might need it, but it won’t make worn out cases new again.

When applied to a specific set of cases that have been all run the same number of load/fire cycles, it can be one of a number of techniques to minimize variability in that brass.  I’m thinking of the way some folks treat that really expensive box of Lapua .308 cases...

On the other hand, if it’s applied to a collection of random, mixed headstamp, mixed load/fire cycle brass, it’s not likely to do anything particularly notable for you.

Now here’s something to consider: people who are diligent in case prep, and who swear by annealing (every few firings or even after every firing) are more likely to start out with high quality brass.  This is a real factor in the equation, and it’s what researchers call a “confound.”  Is the annealing really helping, or is it just that the high quality brass is that good?  I have not ever read a report of someone taking one of those expensive boxes of Lapua cases and separating them into a “anneal” group and a “don’t anneal” group, then comparing the results over a large number of loadings.  I think that there would be at least a measurable difference, even with brass of such legendary quality, but I don’t know that it would be true.

On the other hand, if a batch of LC 5.56 cases just doesn’t seem to group the same way it did when you loaded them as once-fired, it can’t hurt to anneal them and see if they were just getting work hardened.  I HAVE read multiple, well documented reports of this application of annealing actually improving group sizes.

I have this pretty box of lapua brass next to a bag of LC once-fired that I might could be convinced to do just such a test on with my shiny new 223 bolt gun

I never said it doesn't work.  I asked for proof of the claims being made of extended case life and accuracy improvements.  Proof is objective, quantitative data, not words, feeling, hopes or expectations.

My assessment of the benefits of annealing, using my particular criteria, is that a 20% improvement in shots fired is not worth all the cost and labor associated with annealing.

As I said, that case life improvement only happens at the tail end of case life, when most of the rest of the batch has already died. and you lose a significant number of additional cases with each successive firing trying to milk that last bit of case life out of worn out brass (think chamber erosion).

Do you have quantitative accuracy data to back up your accuracy claims?  I can say that both new and old, unannealed brass still shoots as well as I am able (<1/2 MOA).

What I like about the OP's salt bath annealing that it is safe (brass cannot get hotter than the salt) and MUCH lower in cost (~$50).

However, it still adds significantly to labor, may not improve case life as much as one might expect/hope and accuracy claims haven't been verified.

Can you provide a link or the data on this?

Annealing may not hurt the performance of otherwise aging brass but it will hurt your wallet.

If you are thinking of the exercising the annealing option, you have to ask yourself, "Do I want to buy an annealing set up or do I want to chuck this brass and buy some new, possibly premium brass?"

LOL ...I went to annealing because I had proven accuracy loads developed for particular rifles, where I was loosing accuracy on paper as brass reload counts went up = low reload counts compared to high.  Through testing and measurements I narrowed contributors to the necks and /brass top ends.  My measured neck dimensions were loosing consistency (compared to less fired) coming out of the die as reload counts grew, and brass headspace measurement spreads which were spot on with lower recounts, were creeping.   I chose to try annealing to address neck tension, as I suspected inconsistent work hardening was creating inconsistent brass elasticity, contributing to inconsistent spring back coming out of the die (and likely inconsistent grip strength amongst those that did measure the same.)  Annealing addressed the issues....I saw it on my measurements (both necks and headspace measurement consistency), I experienced it on targets (accuracy back to accuracy of proven/lower counts).    This is part of the reason I anneal.

Since then, I have done some other testing.  Through this testing, I have proven to my own satisfaction that annealing incorrectly can do more harm than good in my stuff.  Accuracy performance losses by reloading brass that wasn't annealed beyond production, with underannealed brass, with properly annealed brass, with over annealed brass produced targets worse than knowns all else being as equal as I could get them.    I am never surprised when some people experience accuracy gains with annealing, and some people don't.......so much of reloading is more about the user  than the tools, annealing is no different.

What these experiences, along with what I posted earlier,  taught me is, with my processes and equipment, annealing is beneficial to repeatable accuracy if done correctly, and I am experiencing what I consider to be excellent case life, for me, in my stuff, with my process.....and if I am going to do it, I better consistently get the job done right.  I found other tag along benefits I mentioned in my previous post I won't repeat here ... none of them are shocking to me (or many others who experience the same thing).

While my annealing experience isn't statistically valuable to you (and from a pure stats standpoint, it shouldn't be, nothing I have seen in this thread is), my experiences with annealing are valuable to me and potentially others, which is why I share them... and I am perfectly fine with both (your experience of finding little value in annealing and my experience of finding some).

Experience has taught me there are few absolutes in this game, ymmv

Congrats on your repeated Less Than 1/2 moa with new and old, unannealed brass, wow.

If the salt bath thing (or some other tool) can actually make easier (and in some cases more affordable) for users of disparate levels of annealing experience to produce brass annealed correctly and consistently beyond what we already have, that would be a step up in my book.

Quoted:
Experience has taught me there are few absolutes in this game, ymmv

Experience has taught me that in the arena of reloading and shooting there are all sorts of 'snake oil' for sale backed by 'old wives tales' to facilitate removal of money from your wallet.  Very little or none of the claims are backed by any evidence.

That statement is not directed at you, greenheadcaller.

Congrats on your repeated Less Than 1/2 moa with new and old, unannealed brass, wow.  

It's the rifle and sights, not me and not the brass.  Were I a better shooter, it would likely be better than that.  The rifle is very good.

If the salt bath thing (or some other tool) can actually make easier (and in some cases more affordable) for users of disparate levels of annealing experience to produce brass annealed correctly and consistently beyond what we already have, that would be a step up in my book.  

Agreed, the constant temperature salt bath at the proper temperature differs significantly from a torch's flame.  It should make the process more reliable for a novice.

I'm still awaiting proof of a benefit it but it should be easier and cheaper.  Maybe RocketmanOU will provide some proof in the weeks or months to come.

The thing is, it doesn’t require an expensive machine or even a home made machine.  The budget setup dryflash3 describes is very inexpensive, and it works very well.  The most expensive part of it was the Tempilaq.  I use a cheap deepwell socket, a cheap electric screwdriver, and a nice, yet inexpensive propane torch.  That bottle of Tempilaq will last a very long time.

I can’t provide much in the way of available and accessible quotes for the “multiple reports” of improvement in groups, but I seem to recall first reading about annealing and group sizes in “The ABCs of Reloading” (4th Ed).  And today I found this post in the M14 Forum, and this article on rifleshooter.com (the results are at the end of the article).  It would appear that the best chance to see the effects of annealing on group size would be with a known good load and with carefully prepared brass.

I started a thread for such an experiment. Depending on what we find, I could probably be persuaded to play with the salt bath approach just for the sake of experimental observation, but I want to finish my 650XL and this basic annealing experiment first.

Vietnam was the third of my Father's three wars.  He tells me they used to say something like that to the new guys there, too, except it was DaNang, not Paris.

For the little annealing I do, I just follow the set-up recommended in Hornady Handbook #4 which involves 1) a baking pan, 2) a propane torch and 3) water.

It doesn't get much simpler, or much cheaper than that.

Of course it does require discipline, patience, and focus to make sure each case is exposed to the same portion of the flame for exactly the same amount of time and then is tumbled over into the water at the same time after the flame is removed.  I can do that for 100 cases, but not 1,000, so for many readers the more-forgiving salt-bath annealer may be a good choice.

By the way, this link is now broken.

I have to anneal when making brass for a wildcat round where I chop the parent case at the shoulder junction since the brass was never annealed that low in parent brass form. I use the tempilq, socket and torch method. I noticed where I missed annealing one case and went to resize it then measure its shoulder bump, it was off from spring back. That's how I discovered i missed it since I measure each case upon initial sizing.
I anneaneled it, then sized it again and it came out right.

This is the method I've used in the past but no longer bother.

The 2 articles you link show the problem in the fact there aren't any large scale group testing of this..............

The "M14" article points out the difference is what 0.13"  ??  
Which is worth the effort?  Is that even statistically a difference?

The "Rifle Shooter"  is also flawed depending on how closely you examine the few groups he has.  What I mean is the first 2 groups of the UN-annealed are smaller in size.  It's the third which is bigger and then throws the average over the annealed.  But couldn't I just as easily say 2/3's of the un-annealed groups were smaller?

I saw the melted salt thread a while back as well, and that's what I use now.  Very easy to do.

Worst part was having to buy so much salt to make shipping worthwhile.

Interesting read.

I will say some folks are looking at different aspects of annealing results as a target.

-accuracy through consistency
-case life extension
-facilitating an easier reforming process.

My first dabble with torch annealing was my fight with brass spring back in a Garand.   Fresh once fired brass was consistent in headspace when sized.  After four loadings my shoulder length was dramatically scattered due to the spring back.  I found that totally unacceptable to be adjusting dies or changing Redding match shell holders in an attempt to regain tolerance.  Any way it was an eye opener to see how much spring back could show up.  I never saw it in my bolt guns that bad.

I think for any statistical significance you may need much larger lot numbers.  Maybe even an indoor range with no wind to assess accuracy.   MV might be an easier metric.   At ranges under 300 I can say I have certainly had better SD on velocity deliver less precision that a charge or seating depth with better SD numbers.   I suppose at 800 that would change.

Huh....works for me.

when putting the cursor over the link. its showing a rover.ebay.com redirect for me...

Links to ebay here too.

Me three, however if I right click and choose "open in new window" it will pull up the correct Midway page.