3 component vs. individual milling

[DavidF]

A Basic Comparison of Milling Media and Preparation Methods for Black Powder

By David Forster, May 2017

Background:

I have been experimenting with black powder preparation methods for some time now. Most of my efforts lately have been with powders that were made by 'supermilling' the charcoal, and then combining it with milled potassium nitrate and rubbermakers sulfur. The sulfur was used 'as is' from the bag. With this method, I have been able to match or exceed the performance of Goex commercial black powder. My method takes a little longer than milling the 3 components together, but is useful in situations where running a potentially explosive mill safely is not an option.

With my method, the charcoal is first prepared for milling by running it through a garbage disposer. This reduces it to -10 mesh or so. It is then milled for 6 hours with stainless steel media. The potassium nitrate is technical grade, milled for 2 hours, also with stainless steel media. The charcoal, potassium nitrate, and sulfur are mixed together by hand, and run through a 40 mesh screen twice. The resulting product is analogous to 'mill dust'. It can be used just as one would use mill dust, in any application.

There has been much debate about milling media sizes and types over the years. Many statements are made, but very few test results are available to the fireworker that wants more details before deciding for themselves what is the best way for them to prepare black powder. Most people simply mill the 3 components together with lead media. This method has been in use by amateur pyrotechnists for decades now, with very few accidents reported. It is worth noting that the accidents that are reported are usually not mill explosions. The (very few) accidents mostly occur during handling after milling. Some people mill their black powder with stainless steel media. It is thought that the harder media would grind better. At the same time, stainless steel is less dense than lead. Therefore, the impacts between media have less force behind them. It's possible that these 2 differences cancel each other out. Another reason one might choose stainless steel over lead is because lead can wear down and contaminate the powder. Most feel that this is not a big issue if the lead has been hardened with antimony.

When I promoted my screen-mixing method for making black powder, I was asked if my method could produce black powder of equivalent power to that which was milled as a 3 component mixture in the same amount of time. I was not able to answer that question, since I had made no direct comparisons between methods. My goal was to produce excellent powder with single-component milling, not to win a contest. This question and others have been nagging at me for some time, so I decided to do some more testing. Maybe I would win the contest!

Baseball testing is quite popular among powder-makers for evaluating their powders, and comparing them (hopefully favorably) with the powders made by others. A 10 gram sample of powder is loaded into a 3" (76mm) mortar, and used to launch a standard baseball weighing about 150 grams. A stopwatch is used to measure the total flight time. According to Ned Gorski, host of Fireworking.com, a total flight time of 7.5 seconds will put a 3" shell to a proper display height. I chose this testing method.

The Test Batches

For this series of tests, 6 batches of powder were prepared. Each batch weighed 750 grams. This is a typical batch size for the amateur pyrotechnic hobbyist. To approximate the well-known 75-15-10 formula, I used 563 grams of potassium nitrate, 112 grams of charcoal, and 75 grams of sulfur. The powders for the test pucks (explained later) were drawn from these larger batches. My reasoning here was that I wanted to do things as closely as possible to what I felt other pyrotechnic hobbyists would do.

The powders were milled in Rebel 17 jars, which have a capacity of 5 litres each. The jars have heavy black rubber liners, which dampen the noise- and probably the milling efficiency as well. On my mill, these jars turn at 58 RPM. My mill holds 2 jars. To obtain maximum accuracy, I ran the test batches in pairs, and switched jar positions halfway through the milling time. The total milling time for each batch was 2 hours, 15 minutes. The mill is remotely located and barricaded as any black powder mill should be.

The first pair of tests was to compare the performance of black powders milled as complete mixtures, with the only difference being that one batch used 35 pounds of 1/2" hardened lead spheres, and the other used 30 pounds of 1/2" 304 stainless steel spheres.

The second pair of tests was done to compare screen-mixed powders to each other- and the above pair. In one batch, the charcoal was milled for 2 hours, 15 minutes with the lead media mentioned above. The batch size was 500 grams. In the other 500 gram batch, the charcoal was milled for the same time with the stainless steel media also mentioned above. The potassium nitrate for both these tests was separately milled for 2 hours with 30 pounds of 1/2" stainless steel spheres. The sulfur was used 'as is', straight from the bag. After milling the charcoal and potassium nitrate, the batches for testing were made up by screen-mixing all 3 ingredients together twice through 40 mesh.

The third pair of tests was an afterthought. Since I have various sizes of 304 stainless steel media, I thought I would compare the performances of 2 more powders, milled as a complete mixture, but with smaller media. The first batch used 30 pounds of 3/8" media, and the second batch used 30 pounds of 5/16" media.

At this point I would like to address the issue of lead contamination. Different people have different thoughts about this subject. I wanted to quantify the level of contamination in the batches that used lead media, so it could be considered as a possible variable while reviewing the results I will eventually get to. As mentioned previously, these are hardened spheres, and cast by an experienced man for the purpose of milling black powder. The quality is very good. The powder milled with lead in the first pair of tests gained 5 grams of lead/antimony weight, going from 750 grams to 755 grams. When milling the charcoal for the second pair of test batches, the 500 grams of charcoal milled with lead gained 20 grams. This was ascertained by starting with immaculate mill jars and cleaning them after milling as painstakingly as possible to collect every last bit of powder. The powder that remained on the media and jar surfaces afterwards was insignificant. It is fairly well-known that milling charcoal by itself is very hard on lead milling media. With stainless steel, this issue (if it is one) does not exist.

Preparations:

To achieve maximum accuracy of results, the materials must be uniform to start with. The charcoal I chose for these tests was my own homemade narrowleaf willow charcoal. It was run through a Waste King garbage disposer-based charcoal grinder. This produced a mixture of particle sizes, all the way from 10 mesh down to airfloat. All charcoal used for these tests came from a single bag. Each scoop of charcoal was brought up through the mass from the bottom of the bag, ensuring that I wasn't using just coarse stuff for one test and fine stuff as I got to the bottom of the bag. The potassium nitrate was technical grade, from Norway. The sulfur was rubbermakers sulfur, referred to as H10.

There can be significant variations in black powder performance, based on preparation methods. I chose a method of preparation that is used in the commercial production of black powder to get the most consistent product I could for my comparisons. Commercially, the powder is generally dampened slightly, and then pressed into 'pucks' (or other shapes) using high pressure. I'll just call them pucks. These pucks are quite dense, between 1.7 and 1.8g/cc. The pucks are then mechanically broken up, and then graded for size. At some point they are dried.

I used 6 batches of powder for this series of tests. Each batch weighed 300 grams. The individual powder details will be given later. Each of the batches was spritzed with 8% additional (24 grams) of water. The water was worked in by hand, and the powder was then run twice through a 20 mesh screen. The mixtures were left covered for a couple of hours to 'temper'. Tempering allows the moisture to distribute more evenly throughout the powder, and shouldn't be rushed. After tempering, the powders were briefly mixed again by hand. Each batch of moistened powder was pressed in charges of 51.0 grams. Some liquid was expressed during pressing, which was done at 5 tons of force, using a hydraulic press. The total time to press each puck was about 5 minutes. Each puck lost about 2.5 grams of liquid, and ended up at about 1/2" (12.7mm) thick.

After drying, the pucks were 'corned' and graded. I don't have a black powder factory, so my hillbilly method involved a Louisville Slugger, a piece of PVC pipe, and a test cap. The Slugger was modified, having the end of a meat tenderizer hammer mounted to the tip. The pucks were broken up, and run through a 4 mesh screen, into a box. What remained on the screen was returned to the pipe, given a few more downward hits with the Slugger, and re-screened. This process was repeated numerous times until all 6 pucks had been broken down to pass 4 mesh. After that, the 'powder' was agitated on a 12 mesh screen for at least 2 minutes, to ensure that all small particles were screened out. This 4-12 mesh cut is my homemade version of 2FA. 2FA is a slow black powder (due to the large particle size) used to lift larger shells. About 2/3 of each batch was graded to 2FA. If I were stronger or more aggressive with the Slugger, that amount would have been considerably less.

All batches of pucked, corned, graded powder were catalogued, and stored together for at least a week before official testing began. Only the 2FA fractions were used in this group of tests. The remainders of all powders were further graded to 12-16 mesh to simulate 1Fg commercial powder. Using this grade instead of 2FA adds 1-2 seconds to the total flight time of the baseballs. The -16 mesh powder was saved for other uses.

The Testing Method

For each of the 6 batches, 3 tests were done. Each test used 10.0 grams of 2FA-sized powder grains. A triple beam balance was used to ensure accuracy in weighing. The powders were all contained in identical plastic condiment cups after being weighed out. The cups came with snap on lids, so I punched a small hole in each lid to accommodate a piece of long thin black match, which I taped into the cups. I then snapped the lids on. A short visco leader was taped to the protruding match for lighting.

The launch tube was a standard 3" fiberglass mortar with a weighted base. The mortar was fitted with a bubble level, which could be removed after the mortar was in place and pointing straight up. The baseballs were painted black to make them easier to see in the sky. The individual lidded cups of powder were lowered into place by the black match, and a stick was used to make sure they were seated in the bottom of the launch tube. The baseballs were each lowered into place with a piece of sticky tape, which was yanked off when they were seated on the cups of powder. All tests were done the same afternoon. A standard stopwatch graduated in 1/5 second increments was used to record the total baseball flight times.

Let the Testing Begin!

1) The first powder to be tested was milled as a complete mixture with 1/2" hardened lead media for 2 hours, 15 minutes. These are the flight times: 10.4 seconds, 10.45 seconds, and 10.5 seconds. The average flight time for 75-15-10 milled with 1/2" lead spheres was 10.45 seconds.

2) This complete mixture was milled with 1/2" stainless steel spheres. These are the flight times: 10.1 seconds, 10.2 seconds, and 10.2 seconds. The average flight time for 75-15-10 milled with 1/2" stainless steel spheres was 10.17 seconds.

3) This screen-mixed powder was made with the charcoal that had been pre-milled for 2 hours, 15 minutes with 1/2" hardened lead spheres. The potassium nitrate was milled for 2 hours with 1/2" stainless steel media. These are the flight times: 8.9 seconds, 8.9 seconds, and 9.3 seconds. The average flight time for screen-mixed powder made with lead-milled charcoal was 9.0 seconds.

4) This screen-mixed powder was made with charcoal that had been pre-milled for 2 hours, 15 minutes with 1/2" 304 stainless steel spheres. The potassium nitrate was pre-milled as mentioned above. Here are the flight times: 9.1 seconds, 9.3 seconds, and 9.1 seconds. The average flight time for screen-mixed powder made with stainless steel-milled charcoal was 9.17 seconds.

5) This powder was milled as a complete mix with 3/8" 304 stainless steel spheres. These are the flight times: 9.5 seconds, 9.3 seconds, and 9.7 seconds. The average flight time for 75-15-10 milled with 3/8" stainless steel spheres is 9.5 seconds.

6) This powder was milled as a complete mix with 5/16" 304 stainless steel spheres. The times are: 10.2 seconds, 9.3 seconds, and 9.7 seconds. The average flight time for 75-15-10 milled with 5/16" stainless steel spheres is 9.73 seconds.

Observations, Conjecture, and Conclusions

Well, the results are in. The most popular medium for milling black powder is also the best. The amount of lead contamination per batch of complete black powder milled with hardened lead is negligible, and certainly didn't slow the powder to any significant degree. The stainless steel was so close in performance that the difference isn't even noteworthy. Of course, lead is known to be non-sparking. Stainless steel is harder, and could spark. If hard, gritty contaminants made their way into the mill jar, an ignition would seem much more likely if stainless steel was used to mill complete black powder. Hardened lead is obviously the most prudent choice in this application.

The next pair of results clearly shows 2 things. The first is that in apples to apples comparisons, milling black powder as a complete mixture makes a more powerful product than if the components were milled singly. The second thing is that although the screen-mixed powder is weaker, it is more than acceptable for general use. In this round, the stainless steel beat out the lead, but as above- not by much. The contamination from milling charcoal with lead was certainly significant. The performance was quite good, considering that 4% of the charcoal weight in the 75-15-10 was displaced by the contamination. Even though the powder was good, the flight times were the lowest of all powders tested. I certainly won't continue to mill plain charcoal with any type of lead media. For this application, stainless steel is a better choice.

The last pair of tests was done as an afterthought, since I happen to have full mill jar charges of 2 other sizes of stainless steel media. I thought it would be interesting to see how media size alone would affect the performance of milled black powder. When we compare the results of batches 2, 5, and 6, no obvious relationship seems to exist. The largest (1/2") media gave the longest times, but the smallest media (5/16") didn't give the shortest times. The mid-sized (3/8") media took that position. I have some thoughts on this.

There are a couple more things I'd like to present for consideration. First, there were 3 test flights done with batch #1 that do not appear here. The results were 9.1 seconds, 9.6 seconds, and 10.6 seconds. These were my first tests, and the spread on the times was so great that it did not appear that I was going to derive any valuable conclusions from my baseball testing. After thinking about it, I figured that maybe the spread was due to the fact that I used 3 different (painted) baseballs. The baseballs weighed between 149 and 152 grams. I repeated the tests using a single baseball, and got the results to be much tighter, as reported for batch #1 earlier.

After seeing how important it was to use the same ball for consistent results, I decided to use the same ball for all tests. By the third round of tests, the ball was looking a little battle-scarred. A bit of the stitching had melted away and let go, and a flap of 'leather' was sticking up a little on the ball. The wounds were patched with some carefully laid down masking tape, and the testing continued. The flights of the ball did become somewhat more erratic as the testing went on. It's reasonable to assume that the changes in the ball would affect the aerodynamics and therefore the accuracy of the test results. It's also a known fact that a baseball can be made to curve by the way it is thrown, and by the orientation of the stitching when thrown. It doesn't seem to be much of a stretch to think that the anomaly seen between the results of batches 2, 5, and 6 was due to irregularities in the ball.

A Final Word

While there were no earth-shattering revelations, these tests did prove one thing conclusively. It is not difficult to produce perfectly serviceable black powder using different types and sizes of milling media, whether the powder is milled as a complete mixture, or as single components. All the results in this round of testing were more than satisfactory. The spread between the lowest and highest flight times was fairly close.

The study of black powder has spanned centuries, because it is a complex and important product. It is my hope that this examination has been helpful, if only in a small way, in furthering our general understanding of black powder and its preparation.

-------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

Mumbles, if you want to move this to somewhere more appropriate, feel free. I thought it related 'somewhat' to the topic at hand

[lloyd]

David,

Thanks very much for that in-depth exposure of your experiments. Many of us have been following-along as you've done them, but it's very nice to see it all in summary format.

 

Lloyd

[Mouse]

That was beautiful dave!! I can feel my head grinding for the next week trying to process it all!!

[viziers]

A lot of info to take in and use for everyone.

 

 

vizi

[Mumbles]

I did move this to it's own post. I think it's interesting and engaging enough to warrant it. I love the work by the way.

 

It's a little of topic, and maybe a heavy handed hint, but has anyone ever similarly compared TLUD to retort charcoal in a similar manner?

[Richtee]

That was a fun read!

Nice work!

[PeteyPyro]

Very nice work, David. Thanks for the qualitative testing and for publishing the results for us all to see.

[mabuse00]

Thank you very much for sharing your results with the pyro community David

This is a very valuable piece of work.

 

 

It is worth noting that the accidents that are reported are usually not mill explosions. The (very few) accidents mostly occur during handling after milling

In my case, I still do not like the classic 3 component process...

Especially the dumping of the barrel, after what I read.

 

 

I'd like to mention a third way:

The "Double and double component method".

I has been described on the old Pyroguide page.

 

There are two seperate milling operations, both are said to be incombustible (apart maybe from burning with air in the jars):

-5 parts of charcoal with 75 parts of nitrate

-10 parts of charcoal with 10 parts of sulphur

 

This should be as safe as the individual method, but maybe with a certain performance gain...?

I dont know what actually causes the advantage of the 3 component method, but to a certain extent this might happen with the double & double component method also.

Something like "getting the sulphur (or nitrate) into the pores of the charcoal" is often quoted.

Edited June 3, 2017 by mabuse00

[lloyd]

I'll let Dave espouse his theories, but I want to comment on this, too.

 

I've pretty-much always regarded the 'getting the sulfur (or nitrate) into the pores of the charcoal' as a myth. When the charcoal is as small in particle-size as it needs to be for very fast powder, hardly any of the cellular structure (pores, tubules, etc) continue to exist. Even the cell walls are broken-down into discrete particles.

 

I - personally - don't believe that mechanism even exists, much less that it might be a contributor to powder speed. I think Dave has figured it out correctly -- fine particles of all three components make for the fastest powder.

 

Lloyd

[braddsn]

Mumbles, FWIW, I was cooking my charcoal in a retort for the first 2 years of doing pyro. A year ago, I switched to a TLUD (soooo much more convenient, never looking back) and I still test every batch of bp that I make. I noticed zero difference in speed between retort and TLUD cooking. The only thing that I have found that truly makes a noticeable difference in my bp speed is mill time. I have Ned's mill design, it turns a thumler's jar at 70 rpm's. There is a pretty big difference between 2 and 4 hours. I mill all my bp now for 4 hours. I haven't tried 6 or 8 hours yet.

[uncrichie]

Thank you very much for sharing your results with the pyro community David

This is a very valuable piece of work.

 

In my case, I still do not like the classic 3 component process...

Especially the dumping of the barrel, after what I read.

 

 

I'd like to mention a third way:

The "Double and double component method".

I has been described on the old Pyroguide page.

 

There are two seperate milling operations, both are said to be incombustible (apart maybe from burning with air in the jars):

-5 parts of charcoal with 75 parts of nitrate

-10 parts of charcoal with 10 parts of sulphur

 

This should be as safe as the individual method, but maybe with a certain performance gain...?

I dont know what actually causes the advantage of the 3 component method, but to a certain extent this might happen with the double & double component method also.

Something like "getting the sulphur (or nitrate) into the pores of the charcoal" is often quoted.

The double/double method Mabuse00 mentions was used by the Waltham Abbey Powder Mills in England in the 1800's. They needed a way to make BP and reduce the hazards of a 3 component milling mix. Here's an easy way to remember the milling proportions after weighing. 1/3 of your charcoal is milled with all of the potassium nitrate and the other 2/3 of your charcoal is milled with all of the sulfur. The two results are then screen mixed several times (how many?). I tried it once and had less than stellar results. I blame it on questionable charcoal that may have been mixed hardwoods. I've never tried it with a hotter (Eastern Red Cedar or Paulownia) charcoal. Your mileage may vary.

 

Kurt

[DavidF]

I had less than stellar results with double component milling too. And, I used 2 times through a 100 mesh screen to mix the ingredients!

 

Having since then found that the fineness of the charcoal is a bigger factor than the intimacy of the mixing, I've gone that route. I did mill black powder (and with stainless steel) for this set of tests- just to know what's what. I don't do that for my own powder any more for general use.

 

If I mill the black powder as a complete mix, I would make 2X750=1500 grams in 3 hours, since I run 2 jars at a time. This is starting from crystalline potassium nitrate, and -10 mesh charcoal. So, that's 3000 grams or 6.6 pounds of black powder in 6 hours of mill time. If I take the mill time down to 2 hours instead, I have 4500 grams (9.9 pounds) in that same 6 hours.

 

Now, let's say I mill 500 grams of that same charcoal for 6 hours in one jar, and mill 3 batches (1000 grams each) of the nitrate in the other jar for 2 hours each. I have enough to make 3300 grams (7.25 pounds) of complete black powder in 6 hours of total mill time. This powder will be at least equivalent to the complete-milled product mentioned above.

 

The tests I did above had nothing to do with finding out what is 'the better way' for me. I'm already satisfied with my way I just wanted to approach the subject of screen-mixed powder from yet another angle. I will do more series of test in the future. I like learning new things, probably more than I like making stuff. And, as usual: thanks for the kind words folks.

[uncrichie]

Dave, I'm looking forward to trying your method. You certainly have put a lot of time/research into this with what looks like excellent results. I'll let you know how it turns out. This time I'll be using known Red Eastern Cedar charcoal!

 

Kurt

[mabuse00]

 

I had less than stellar results with double component milling too. And, I used 2 times through a 100 mesh screen to mix the ingredients!

Did you ever directly compare this to your individually milled powder?

 

 

 

[DavidF]

mabuse00, I did not do much work in that area, and I was using nozzleless rockets and the Acme test stand to evaluate my powders at that time. Perhaps I shouldn't have commented, based on the little information I was able to call up on it. I didn't record much detail on the process. Milling ERC charcoal 50/50 with the potassium nitrate, and then mixing with more (milled) nitrate and sulfur gave an excellent result.

 

I did a test where I ignited the milled 50/50 mix in a closed mill jar with ceramic media, just to see what would happen. It's here:

 

https://www.youtube.com/watch?v=AhIFNm3vutA

 

It was more of a bawoof than a boom It didn't seem that much safer than milling a complete mix, and the screening was a PITA, so I moved on.

[Mitchell]

So it's been there years now and i wanted to know if single component is still the way you prefer toake bp or if you had started doing some other way?

[justvisiting]

Well, that's a blast from the past. Yes, I still do single component milling, and there are several others now It may not have become the preferred way to make black powder for most folks, and that's perfectly understandable. It's gratifying enough that the myth that single component milling makes inferior powder has been debunked.

 

I will probably mill the sulfur and charcoal together now because I ran out of the superfine rubbermakers sulfur and ended up buying prilled sulfur.

[Mitchell]

So is milling the sulfur with the charcoal beneficial? I have seen it both ways throughout the forums. What's you're opinion and why?

[justvisiting]

I'm sure it's beneficial, but how beneficial, I don't know personally. My impression is that the particle size of the sulfur isn't nearly as important as the particle size of the charcoal. But, then there's the intimacy of incorporation. I should look into the question more.

[Bourbon]

Would you still opt for SS media over lead to mill sulfur and charcoal together?

[justvisiting]

Yes, I would.

[Arthur]

The interesting thing to me from the OP's project is that he pressed and corned his samples in the traditional way.

 

Many powder makers do not press and corn their product but do grate or rice a large moist lump. It's my opinion that pressing and corning helps make better performing powder as part of the incorporation of the ingredients.

[Mitchell]

Arthur in your opinion is pressing and Corning better? Is this because dextrin isn't present generally? And i agree about SS media as well. I'm switching because i don't want to deal with lead contamination

[justvisiting]

Arthur, I pressed and corned the powders to limit the effects of external variables, so apples to apples comparisons could be made. I'm sure pressing the damp mixture into pucks gives better intimacy of incorporation than simply screen-mixing and granulating, as you suggest.

 

Mitchell, I don't think pressing and corning is better- or not better. Black powder has many different uses, and it's certainly better for some of them. Pressing, corning, and grading gives powder that has excellent durability and consistency from batch to batch. It's a lot of labor though. Lots of folks simply screen-granulate their BP, as Arthur mentioned. It's important to note that screen-mixing and screen-granulating are two different things. For instance, ball milled black powder is often wetted with solvents like alcohols (or water) and granulated. Or, the same powder might be milled with 5% dextrin and coated onto rice hulls to be used as shell burst. Screen-granulated black powder does 'better' than pucked, corned powder in baseball tests by 1- 1 1/2 seconds. For nozzleless or end burner-type rockets, the requirements for the powder are different yet again.

 

I think one thing that trips up newer pyros is the vast amount of seemingly contradictory information about black powder.

[Mitchell]

justvisiting, got it thanks. I have up to this point just screen granulated but had of course seen making pucks and corning around and just wanted to ask if there was an inherent "advantage" and your reply was exactly what I was looking for. You answered my question, I appreciate it.

 

And yes, when I started a few years back I was definitely overwhelmed with all of the information out there, and all the different techniques there is to make everything. My advice to my former self is find a solid way to make bp and just keep doing it until the results achieved are consistent every single time. I spent way too much time with inconsistent results.