Salt Bath Annealing setup and quantitative assessment.

[entoptics]

Just finished constructing and first testing of a salt bath annealing system. First post will just be the components and schematics for the rig I built. Second post on opinions of use and efficacy testing.

Links are amazon, cause they're easy to do. Shop for yourself. Prices are estimates for OTD after tax/shipping/etc. Comments regard "musts" if substituting other items.

Temp Controlled melter setup...

Lee Melter - $40
PID Controller - $30 Make sure it's Solid State Relay (SSR) compatible.
Thermocouple - $12 Make sure it's rated for above 500° C
Solid State Relay and heatsink - $15 Make sure it's rated for ≥ 20 amps.

The salt is a near eutectic 60/40 by weight mixture of KNO3 (potassium nitrate, saltpeter) and NaNO3 (sodium nitrate). The mixture isn't critical and my research indicates anywhere from 60/40 to 40/60 will work. I scrounged mine up for free, but they are also commonly available online and perhaps even locally at a farm supply type place. You need approximately 340 grams (12 oz) to fill the pot. Unfortunately, they tend to be sold in 1-2 lb bags, so you'll end up spending $30-$40 to get 2-4 times what you actually need for a lifetime of annealing.

Other items...

Wall outlet - $1
Wall outlet cover - $1
Light Switch- $1
Light Switch cover - $1
Mounting board - Free to $10
12-16 gauge electrical wire - Free to $1
Outlet/Switch housings - $5

Shell Holder...
2 each ~5"x5" scrap steel or aluminum sheet
Long bolt and 3 nuts - $1

As you can see, it's about a $120 total investment.

Schematic for my personal assembly. Be sure not to place the two housings too close together, or you won't be able to install the cover on one of them. Only a complete ******* would do something like that though…

Template for the shell holder I built is attached as a printable PDF below, which you can use to cut your plate and/or center punch the holes. It should print exactly to scale. The top plate rests on top of the melter, and indexes on the 3 screws on top of the pot, and the mouth of the case sits on the bottom plate. These are joined by a 5/16 bolt, with a nut securing the bottom plate, and a nut on each side of the top plate. This allows you to precisely adjust the depth the cases will sit in the salt.

I chose 223, 308, and 300 WM for hole sizes. You'll have to do your own math for what size brass you intend to use.

So...Now your thinking..."But AMP says it doesn't work!". That'll be addressed in the next post...

 

[entoptics]

To quantitatively assess the efficacy of annealing with the salt system, I took 40 pieces of well used brass, annealed half for 6 seconds at 500° C ± 5° C (932° F ± 9.3° F), and left the other half alone. All brass were treated the same except annealing, and all were loaded on the same equipment in the same session.

Brass Specifications
MAI brass (Norma manufactured)
≥4 firings
Originally neck turned
Weight sorted to ± 0.2 gr
Length sorted to ± 0.002 OAL
FL Sized to bump shoulder 0.002" ± 0.001"
Bushing sized to ~0.002" neck tension

Load Specifications
208 ELD-M
76.2 gr H1000 hand weighed to exactly the same charge on PACT scale
Federal 210M primers
≤ 0.003" runout (most <0.002")
3.590" OAL (~0.02" off lands)

Shooting Specifications
Savage LRH 300 WM with 24" Shilen Heavy Sporter bbl (otherwise OEM)
5-25 Sig Sauer Whiskey 5 at 25X
200 yard target
65° F, 55% RH
Obnoxious, but mostly bearable, mirage
2-5 mph full value wind, left to right. I tried to shoot in the breaks (makeshift flags at 100 and 200).
Prone, bipod, rear bag rest, chest/chin in a makeshift "nest" of pillows and blankets. With this setup, I can routinely hold the crosshair well inside a 0.25 MOA circle and break the trigger.

Testing Results
Red = Not annealed (n = 19)
Blue = Annealed (n = 19)
4 shot groups were fired, alternating from annealed to un-annealed for a total of 40 rounds.
Velocity was measured using a LabRadar. Individual tracking files were inspected and culled to reduce errors associated with LabRadar's wonky algorithm for calculating velocity, which is subject to spurious readings from one or two bad tracking data points.
Groups were measured using SubMOA shooting app for the iPhone.

Velocity

Group Size

Summary of results

Annealed brass (n = 19)
Average velocity = 2783 fps
Extreme spread = 40 fps
Standard Deviation = 9 fps
Average group size = 0.55 MOA (0.36 MOA minus worst flier)
Smallest group = 0.43 MOA
Largest group = 0.74 MOA

Un-annealed brass
Average velocity = 2797 fps
Extreme Spread = 60 fps
Standard Deviation = 16 fps
Average group size = 0.74 MOA (0.39 MOA minus worst flier)
Smallest group size = 0.51 MOA
Largest group size = 1.07 MOA

Comments on results
I scrubbed the crap out of my rifle before I did the tests, which might explain the gradual velocity loss through time. The first two groups were also terrible, so I omitted them from the results, based on the same logic. The ES/SD for both sets is biased a little high because of the velocity drift. In all cases, the individual 4 shot strings had lower ES/SD than the average.

Summary
Though of limited scope, I feel that this test indicates that salt bath annealing has a positive affect on precision. This was literally my first attempt at annealing, using the "wisdom of the crowds" (i.e. google) recommended setup and parameters, and I saw nearly a quarter MOA better groups and ES/SD was reduced by ~30%.

 

[ppoole]

Well done it obviously brings some level of repeatability to neck tension etc. Thank you, I'm still curious about hardness, but contempt about precision.

 

[orange1948]

Well done. Thanks.
Duane

 

[Nickie.01]

To quantitatively assess the efficacy of annealing with the salt system, I took 40 pieces of well used brass, annealed half for 6 seconds at 500° C ± 5° C (932° F ± 9.3° F), and left the other half alone. All brass were treated the same except annealing, and all were loaded on the same equipment in the same session.

Brass Specifications
MAI brass (Norma manufactured)
≥4 firings
Originally neck turned
Weight sorted to ± 0.2 gr
Length sorted to ± 0.002 OAL
FL Sized to bump shoulder 0.002" ± 0.001"
Bushing sized to ~0.002" neck tension

Load Specifications
208 ELD-M
76.2 gr H1000 hand weighed to exactly the same charge on PACT scale
Federal 210M primers
≤ 0.003" runout (most <0.002")
3.590" OAL (~0.02" off lands)

Shooting Specifications
Savage LRH 300 WM with 24" Shilen Heavy Sporter bbl (otherwise OEM)
5-25 Sig Sauer Whiskey 5 at 25X
200 yard target
65° F, 55% RH
Obnoxious, but mostly bearable, mirage
2-5 mph full value wind, left to right. I tried to shoot in the breaks (makeshift flags at 100 and 200).
Prone, bipod, rear bag rest, chest/chin in a makeshift "nest" of pillows and blankets. With this setup, I can routinely hold the crosshair well inside a 0.25 MOA circle and break the trigger.

Testing Results
Red = Not annealed (n = 19)
Blue = Annealed (n = 19)
4 shot groups were fired, alternating from annealed to un-annealed for a total of 40 rounds.
Velocity was measured using a LabRadar. Individual tracking files were inspected and culled to reduce errors associated with LabRadar's wonky algorithm for calculating velocity, which is subject to spurious readings from one or two bad tracking data points.
Groups were measured using SubMOA shooting app for the iPhone.

Velocity
View attachment 135408

Group Size
View attachment 135409

Summary of results

Annealed brass (n = 19)
Average velocity = 2783 fps
Extreme spread = 40 fps
Standard Deviation = 9 fps
Average group size = 0.55 MOA (0.36 MOA minus worst flier)
Smallest group = 0.43 MOA
Largest group = 0.74 MOA

Un-annealed brass
Average velocity = 2797 fps
Extreme Spread = 60 fps
Standard Deviation = 16 fps
Average group size = 0.74 MOA (0.39 MOA minus worst flier)
Smallest group size = 0.51 MOA
Largest group size = 1.07 MOA

Comments on results
I scrubbed the crap out of my rifle before I did the tests, which might explain the gradual velocity loss through time. The first two groups were also terrible, so I omitted them from the results, based on the same logic. The ES/SD for both sets is biased a little high because of the velocity drift. In all cases, the individual 4 shot strings had lower ES/SD than the average.

Summary
Though of limited scope, I feel that this test indicates that salt bath annealing has a positive affect on precision. This was literally my first attempt at annealing, using the "wisdom of the crowds" (i.e. google) recommended setup and parameters, and I saw nearly a quarter MOA better groups and ES/SD was reduced by ~30%.

Thank you for your time and attention to detail to help other shooters, especially those on a budget, decide if there is benefit to salt bath annealing. Very good shooting too!

 

[epoletna]

Thanks for doing this, and more importantly (to us) for publishing the results. Especially after the recent bashing of salt bath annealing. Interestingly enough, on your plot of speeds, it appears the annealed brass gave consistently slower velocities, but with smaller deviation.

 

[entoptics]

Thanks for doing this, and more importantly (to us) for publishing the results. Especially after the recent bashing of salt bath annealing. Interestingly enough, on your plot of speeds, it appears the annealed brass gave consistently slower velocities, but with smaller deviation.

Glad to be of help.

The annealed brass definitely went slower. I meant to mention that result in the initial post. I think this alone proves that something is happening to the brass during the annealing, which contradicts the results AMP published.

Based on bullet seating feel, before even going to the range, I suspected there would be a velocity loss. The annealed brass seemed to take ever so slightly less force, and seemed to be a little more smooth and consistent as the bullet went in.

I attribute the lost velocity to slightly less neck tension, therefore a little less pressure build before the bullet is released.

These loads were pretty wimpy, as I used an old recipe developed with a different lot of H1000 and newer brass. I'm probably gonna do another identical experiment, but aim for another 100-120 fps. It will be interesting to see if the same trend occurs with a substantially friskier load.

 

[epoletna]

Hmmm. I wonder, given the feel of the bullet seating in the annealed brass, if it would be ok to step down one or two thousandths in neck bushings after annealing? Any thoughts on that?

 

[epoletna]

And just in case I can score the same source, where did you find the free KNO3 and NaNO3 ? I mean is it in a garden store or something?

 

[entoptics]

Hmmm. I wonder, given the feel of the bullet seating in the annealed brass, if it would be ok to step down one or two thousandths in neck bushings after annealing? Any thoughts on that?

I'm considering that very idea. My guess, based on how subtle the "feel" difference was, going 0.001" tighter would be more than sufficient. First I'm gonna keep testing though. If an extra couple tenths of a grain of fuel will get the velocity back up, then I likely won't mess with the bushing.

And just in case I can score the same source, where did you find the free KNO3 and NaNO3 ? I mean is it in a garden store or something?

I got it from some ancient lab supplies I had lying around at work. The KNO3 should be easily found at a farm/garden type place. The NaNO3 might be the harder one to find.

I also found this person selling the stuff on Etsy. Still a bit pricey with shipping, but not that bad for good supply.

 

[epoletna]

Thanks for the link to Etsy. I've ordered that.

 

[entoptics]

Thanks for the link to Etsy. I've ordered that.

To conserve some of the salt, I set a "drip pan" next to the pot. I shake each case into the drip pan before dropping it into the water. It hardens instantly, and can be flaked off the bowl and put back in the pot. There's a surprising amount rescued.

Just make sure your pan can handle VERY hot stuff, and has relatively smooth surface so the salt doesn't stick to it. I used a medium sized pyrex glass bowl.

 

[epoletna]

To conserve some of the salt, I set a "drip pan" next to the pot. I shake each case into the drip pan before dropping it into the water. It hardens instantly, and can be flaked off the bowl and put back in the pot. There's a surprising amount rescued.

Just make sure your pan can handle VERY hot stuff, and has relatively smooth surface so the salt doesn't stick to it. I used a medium sized pyrex glass bowl.

How many cases do you "dip" at a time? Any reason not to make all six holes the same diameter (for a specific) case, and dip six at a time? Fred

 

[entoptics]

If your on FB we would like to invite you to 2 groups focused on this very thing[Brass Annealing Anonymous and Salt Bath Annealers] and we too have been having seriouos debates and looking over info. To me its about the real world results nto necessarily how it test in a lab. If we can prove we get better SD/ES and brass does last longer that is all we need.

I appreciate the offer, though I'm not on Facebook. I'd suggest that the real world can be a lab if one is careful. Good luck with your experiments. I'm sure they'd be appreciated if you post them on the forum.

How many cases do you "dip" at a time? Any reason not to make all six holes the same diameter (for a specific) case, and dip six at a time? Fred

Having two irons in the fire is tricky enough for me. With only 6 seconds between drops, there's barely enough time to shake off one and reload another. One could probably get a routine down, but if you're not careful, you could ruin brass, or worse upset the pot or otherwise splash molten salt around.