Induction brass annealer redux

[itchyTF]

I’m sure others will chime in, but I prefer more headroom. I built one for a friend (not the MGNZ) using a 36v 1000w. No problems.

 

[oliverpsmile]

Because the neck temperature is climbing fast in this context, it will climb quickly from 950F to over 1000F and beyond. So, when you pause for a heartbeat at The Draper Point for a two count, you get that Goldilocks result but not because you had it at exactly 950 or 1000. Neck annealing has temperatures that are a moving value, not ones that have lots of dwell at some steady equilibrium.

Thank you for your professional comments on the matter.

Just a note:

Two members of this forum manage to stop the temperature climbing at the Draper Point or 1000F (thempilaque), using PID or discrete feedback control systems. Holding the neck temperature for a reasonable time ( 4-5 seconds or more, well above the disputed Reese 1.88 sec) provides some temperature rise at the shoulder of the case. The annealing level (so to speak) does not change during the prolonged time. Some articles published in this forum showed that. The questioned accuracy of the sensors, in this case is irrelevant, as long as it is constant.

 

[hdmunger]

...and consistent.

 

[RegionRat]

Thank you for your professional comments on the matter.

Just a note:

Two members of this forum manage to stop the temperature climbing at the Draper Point or 1000F (thempilaque), using PID or discrete feedback control systems. Holding the neck temperature for a reasonable time ( 4-5 seconds or more, well above the disputed Reese 1.88 sec) provides some temperature rise at the shoulder of the case. The annealing level (so to speak) does not change during the prolonged time. Some articles published in this forum showed that. The questioned accuracy of the sensors, in this case is irrelevant, as long as it is constant.

No need to debate if their systems are or are not really at 1000F or that they are stable.

I am happy for them and encourage their innovation and experimentation. If it works for them, I am all for it regardless of how they got there.

We don't need to get too critical of Reese's 1.88 seconds since we know it is in the right ballpark when his power level is in that league and his temps are obviously transient.

If folks want to go faster or slower, they are free to do so and can use hardness or seating force to see if they got where they wanted to go.

Production lines have produced cartridge brass without closed loops on radiometric temperature sensors on the necks for over 100 years. The hardness stays well within goals. There are good reasons to keep things simple when possible.

If I were still in charge and my juniors asked about changing this, it would take some compelling reasons that are very unlikely to exist in a real-world plant, but I would listen to their pitch and then decide.

Let me put this another way to explain why I am not pushing for radiometric closed loop versus just having a stable heat-rate and marking time from the Draper point. I would ask two questions; will the technology change hold a better hardness tolerance and will the reliability of the process be improved?

If the annealing process is an inductive system, the reliability of the process is still based on the sum of all the parts. It isn't impossible to get a higher reliability with more parts, but it is very difficult in this context.

Now, in my career I have gone the way of placing complicated closed loops on things, and in other places I have wanted things as basic as a rock. I was asked to prove those choices were based on performance and reliability, not that cost doesn't come up as well.

My goal was to get our kids home alive, and that reliability was often asked to be what is called "Class A" which translates to national infrastructure having less than 5 failures in 10,000 of whatever they say in the contract.

If we look at Class A Requirements, that means higher part counts that can fail are not going to make the design review easier. In so many words, if my job is to keep that production line operating without unscheduled maintenance, and to prevent defects from escaping, then the design needs to be more reliable and the vast majority of the time that means less complex not more.

Once we have a steady heat rate in the process (flame or induction) and know how long to pass the brass through it, the hardness variation within the piece-to-piece variation in the brass can be set-it-and-forget-it.

It would be different if each batch required every piece to have a custom time or power level. Brass that is that different shouldn't be part of a "batch" anyway. So then why solve the loop control problem over and over again when a simple time and power setting will do?

When induction-based annealing fails, it is usually because some parts of the power electronics run hot and the materials break down. I generally wanted those RF generators to stay as cool as possible and transients were not their friend. Steady state operation with optimal balance and cooling was the key. Now if I added a power loop, those requirements would go up and the reliability would go down just because of the higher part count, the complexity, and the heat cycles.

I can almost guarantee the technology would cost more because it has more electronics and contains sensors that require complex calibration or maintenance. The temp loop is solving a problem on each piece that can be set once for the whole batch. Adding the risk of more complex electronics to make a correct measurement and control the power on each piece is lowering reliability in general and does it really produce any improvement? Does that make sense?

Stopping a line to swap out and/or calibrate a timer during a batch stop is one thing. Having a line go down unexpectedly because an infrared sensor loop goes wonky or an RF generator fails is another. I am not saying those technology improvements will never happen, they just haven't happened yet.

What experiments and inventions cottage industry folks create has to make sense for their context. I am only trying to quell the fears folks have when they don't have access to industrial metallurgical labs.

I don't want that to mean folks should stop trying to innovate high performance temperature feedback loops for things like this, but we also don't need to overly complicate the concept of good home annealing of cartridge necks either. Alpha brass happens to come with its own built in temperature indicator. We should be happy about it.

At the same time, I have gone on record saying that 750 Tempilaq wasn't very helpful in this context, but the 400F in the early learning curve can be.

It shows how far down the body that heat level will propagate in context. My only goal is to substitute the concept of that 750 Tempilaq with the Draper Point and bow out. I really don't mean to go this far down the rabbit hole, but also don't want folks to be misled about the metallurgy.

We both know the temp measurement accuracy is not irrelevant, but I understand the spirit of what you are saying and what I stated isn't in conflict with that concept. There is a range of times and temperatures that will allow the material to change and reduce hardness to the range we want. After all, some folks don't anneal at all and make that work too.

The technology to "watch" grain growth in real time on case necks doesn't exist, but if it did we would see that thin alpha phase brass isn't hitting some critical phase-change event at these temperatures, so the process is forgiving to an extent.

The heat treatment can be done using that Draper Point in a time level that will work for hobbyists just like it does at the factory and that is all that matters in my point. I don't want to stand in the way of folks innovating their induction machines by any means. The Annie is designed and built very different than industrial lines, so I don't have a dog in the fight.

Cartridge brass is Goldilocks for our purpose. It even gives us a built-in temp sensor. I wish the folks experimenting with temp loops on their induction rigs all the best and will cheer them on. No worries.

Merry Christmas and Happy New Year!!!

 

[itchyTF]

It even gives us a built-in temp sensor.

Not sure what this means.

 

[RegionRat]

Not sure what this means.

That means that cartridge brass will hit the Draper Point and begin an infrared/visible glow at a temperature that by luck is a good indicator for annealing.

You don't need to buy any sensors, Tempilaq, etc. You can just use your eyes and a stopwatch because this temperature indication is serendipitously built into alpha brass.

 

[itchyTF]

I built an induction annealer about 10 years ago. I designed the control to be by time with .01 sec resolution. When the flame sensor was all the rage I thought I'd incorporate that too. I can select either mode. I did not incorporate a closed loop where I can maintain temperature for a select time. My question (probably been touched on already) is since the induction method is typically very short, how much higher than the Draper point would the brass need to go to be annealed? Or, is it simply determined by trial & error?
It would be nice if it was something like if Draper was hit at 2.1 seconds then 50msec more would be a good target. Probably not that simple.

 

[oliverpsmile]

You don't need to buy any sensors, Tempilaq, etc. You can just use your eyes and a stopwatch because this temperature indication is serendipitously built into alpha brass.

@itchyTF to avoid endless trial and error for every caliber or batch of brass, creating data base like AMP, here is a simple approach. Yes, it is not applicable for a production line. But, on our hobbyist level it works remarkably well.

See note above. At this point, using some simple micro controller (Arduino UNO for example), record the reading of you sensor. The reading is your numerical Draper Point. From there make your program starts checking the the sensor reading every few milliseconds (few is determined by the switching capabilities of your SSR). If the reading is +/- 1%, switch the SSR ON/OFF (i.e. it works like a simple house thermostat). You can prolong the annealing time of your liking. Just avoid overheating the brass body. I stop when I got Lapua new brass color at 1/4 inch below the shoulder.

As long as you keep same neck and sensor positions, there is no need to make any changes of your program.

 

[RegionRat]

Or, is it simply determined by trial & error?
It would be nice if it was something like if Draper was hit at 2.1 seconds then 50msec more would be a good target. Probably not that simple.

It really is that simple as long as you understand that with this method (or even with a closed-loop method), you never really know if your process is "right" till you either check the seating force or measure the hardness. YMMV

 

[oliverpsmile]

So, then, it's time to start developing means to check the seating force - electronically of course..
Is there any acceptable, in your opinion, criteria for that? I saw one built by AMP, but the owner did not know what it was good for.

 

[RegionRat]

So, then, it's time to start developing means to check the seating force - electronically of course..
Is there any acceptable, in your opinion, criteria for that? I saw one built by AMP, but the owner did not know what it was good for.

Not sure I understand what you meant by that last line, but the AMP Press is the easy button.

It isn't for everyone and isn't as cheap as a force pack, but when folks use my AMP Press they are usually enlightened.

Some folks go as far as to create their own from scratch, and those work too.

In our labs, we built many different kinds of adaptions for Instron and MTS systems (typical tensile test systems), as well as building many custom versions from scratch for special dedicated test purposes.

With these kinds of systems, the annealing-effect study is easy since the data is plotted in terms of force versus displacement or force versus time during programmed speed. The data is a two axis plot without the need to use your Mark 1 Eyeball to watch a gage while you pull a lever.

With a typical force pack on a Wilson Arbor press, you can still learn what your annealing is doing to your necks but it takes some skill to watch the needle as you operate the press. You end up mentally capturing the difference between the starting effects versus the rest of the seating effects. Many folks have been loading this way for decades without the benefit of an analog force v displacement plot. If your budget won't tolerate an AMP Press, a Wilson force pack is a decent alternative.
ETA: I forgot to mention the 21st Century version that uses hydro-pressure reading on the press as well.

 

[oliverpsmile]

With these kinds of systems, the annealing-effect study is easy since the data is plotted in terms of force versus displacement or force versus time during programmed speed. The data is a two axis plot without the need to use your Mark 1 Eyeball to watch a gage while you pull a lever.

That's it, what I was looking for - the two axis plots vs Mark 1 Eyeball. I'm using K&M arbor press with a gauge and trying to sort the reloads based on the initial seating force.

Thanks

 

[RegionRat]

Yes, the plots are much easier to understand....

 

[hdmunger]

Now someone needs to design, build, and share a simple way to obtain that graph for those who can't, or don't want, to buy an AMP press...takers?

Its certainly not me, but I might be interested in building one if the parts were reasonably priced...

 

[oliverpsmile]

Now someone needs to design, build, and share a simple way to obtain that graph for those who can't, or don't want, to buy an AMP press...takers?

Its certainly not me, but I might be interested in building one if the parts were reasonably priced...

As long as you help with the spelling

 

[hdmunger]

I'm sorry...wasn't correcting your spelling, just amplifying your statement.
To me, constant means it works every time, and consistent means it works the same every time. I didn't mean to split hairs...

 

[oliverpsmile]

No sweat.

 

[chalkeye]

I remade the coil per MGNZ specs, and it works! So that particular mystery appears solved.

I am successfully heating brass but finding I need very long duration (6.5-7 seconds) to get the neck of my BRA cases just faintly glowing in a dark room.

The neck of the case is sitting dead centre of the coil vertically and concentrically.

Any thoughts? have other heights proven better or worse for you guys, or do you think something else is wrong?

 

[itchyTF]

I remade the coil per MGNZ specs, and it works! So that particular mystery appears solved.

I am successfully heating brass but finding I need very long duration (6.5-7 seconds) to get the neck of my BRA cases just faintly glowing in a dark room.

The neck of the case is sitting dead centre of the coil vertically and concentrically.

Any thoughts? have other heights proven better or worse for you guys, or do you think something else is wrong?

What did you do differently with the coil? How much current is it drawing with the case in the coil?

 

[chalkeye]

What did you do differently with the coil? How much current is it drawing with the case in the coil?

Empty coil 8.7A max at 47v
With brass 11.8A max