I have always enjoyed the Passfire articles where they take a commercial shell, carefully disassemble it, and weigh or otherwise analyze the shell contents. Much can be learned by the amateur by doing this carefully, especially such things as ratios... lift to shell weight, burst weight (and type) to star weight, number and size of stars, pasting methods, etc. Further, in the last few months, I have been collecting reagents and the associated techniques to determine the compositions themselves, especially nitrate vs. perchlorate vs chlorate.

I started with one of my favorite 1.4g devices, the ubiquitous 1.5" reloadable mortar. These pack a nice bang for the buck, and by the case, cost less than a buck apiece. Further, they can be easily modified to act as inserts by removing the lift charge. As a bonus, you get some decent lift BP and a length of fast visco that can be recycled.

Since this was posted as a thread, I'm going to be a lazy ass and do a bit of cut and paste.

Here in Texas, we have a gigantic 1.4g consumer works stand nearby, and every 4th, our family drops by to pick up consumer fireworks to fill out what little I have made. My personal favorites are the super-cheap 1.5" reloadable mortars. Today I bought a case (12 tubes, 72 shells) for $64. Driving home, I was thinking about the cost, the bang for your buck, if you will. Don't ask me why, but I decided to dissect one of these shells, weigh the materials, and figure out what it would cost ME to duplicate one of them.

The shell was a 1.5" pasted paper spherical shell, brocade blue. The lift cup was a short section of light cardboard tubing, with paper glued on the far end, and pasted securely to the shell itself. Separating the lift from the shell with a razor blade revealed 5.8 grams of what appeared to be a good quality BP. This BP had a high density and excellent grain strength, obviously pressed and corned.

The shell itself, minus the lift and cup, weighed 26.7 grams. Inside the shell was 4.7 grams of lightly coated rice hull burst, probably 3:1 or 4:1, and 12.4 grams of well-made, excellent small round stars.

For the cost analysis, I am going to exclude the cost of the kraft paper, glue, rice, mortar tube, and packaging. Chemical prices are average, mostly from Skylighter, and labor is not considered. To duplicate this shell at home would require the following materials, assuming the burst is black powder. As we shall see, that is not the case, but for now, bear with me.

Total weight of BP in the shell: 9.56 grams

Breaking down the constituents in the classic 75:15:10 ratios, and using store-bought airfloat charcoal, I calculate the following costs:

KNO3 = $1.50 per pound

Sulfur = $4.00 per pound

Charcoal = $4.00 per pound

A bit of math reveals that the 9.56 grams of homemade BP cots a whopping 4.5 cents! Nice and cheap.

The stars... I am going to consider them 100% potassium perchlorate. The components of a star vary wildly, but the bulk is potassium perchlorate, and most other star chemicals fall into that cost range, or very close to it. At $9/lb for skylighter perchlorate, the cost:

stars = $0.27.

Ignition: These shells use a form of fast visco rather than QM. Fast Visco is available from Cannonfuse, among other sources. It is very effective and handy, but not particularly cheap. The cost for fast visco from Cannonfuse is $0.20 per foot, and the length of the fuse on the shell is about 1 foot, so we'll call it:

Fast Visco = $0.20

Time fuse: You can adapt visco to act as a time fuse, and in fact, it is economical to do so for small devices, but if you are working on a 3" or larger shell, anything less than real time fuse is false economy, IMO. For this argument, we'll use real time fuse at...

Time Fuse = $0.12

Hemi: A hemi set is probably the most expensive part of these devices. The majority that I found were 50 cents or so. Too much. I did find a set of 1.5" paper hemis, though, for only...

1.5" paper hemi set = $0.20

Adding up these components (and I know a lot of the stuff can be found cheaper):

Hemi: 0.20

BP: 0.045

Stars: 0.27

Time fuse: 0.12

ignition: 0.20

Grand total: 83.5 cents

My cost in a case of 12 boxes, 12 tubes, and 72 shells was $64, or 89 cents per shell. When you consider the labor that went into this shell, it's a pretty good deal. Obviously we are all here to make better fireworks than you can buy, and there are a hundred sites where you can discuss consumer fireworks. There's an obvious joy in DIY and you can craft shells you cannot buy anywhere without a license. I don't know why I did this other than pure, dumb curiosity. I have always enjoyed the value of consumer fireworks - they are a lot of fun for the buck, and here we have a perfect example. It is interesting that the costs were so close. The Chinese factories buy in a bulk that we cannot dream of, allowing them to turn a profit. The Chinese ladies who paste these shells probably can paste one up in 30 seconds. It would take me a half hour to do it as well.

The biggest surprise to me was the lift (excellent BP) and the beautiful little stars that look like they were moulded with a pill machine. They each are flawless, identical spheres.

Yankie correctly surmised that the burst was probably NOT BP which would perform poorly in so small a shell, so the decision was made to test the burst and see what it was truly made of.

Visually, the burst has a thinner, sparser look than 5 or 6 to 1 BP on rice. If I were to guess, I'd say it's more like 3:1 or less. To test, I needed to separate the water-soluble oxidizer from the rest of the burst. This is easily done by soaking it in hot water for a few minutes.

This was filtered into a small flask:

The water + oxidizer was added to two test tubes in small proportions. The middle test tube was tested for the presence of chlorate using fenamic acid + sulfuric acid, the now well-proven NPAA test. It came out negative. No chlorate. The next test (left test tube) was a larger sample portion, using methylene blue to test for perchlorates. In the presence of perchlorate, methylene blue forms a distinct, purple precipitate. Immediately, upon seeing the results, I thought "perc" but just to be sure, I added methylene blue to a solution of potassium nitrate, as found in BP (right test tube) to compare.

The results, from left to right:

Methylene blue perchlorate test: Positive. NPAA chlorate test: Negative. Potassium nitrate + methylene blue as a comparison. Note the nitrate blue color vs. the perchlorate purple, plus precipitate.

Conclusion: The burst in this 1.5" shell is KP, probably 2 or 3 to 1, on rice hulls.

For the future, I plan on cutting a few more open, different brands, etc. It is both interesting, and as they say, educational to boot.

One thing I still need to do is count and measure the perfect spherical stars, and test a couple of them for chlorate as well. I fully expect them to be chlorate free, and be predominantly perchlorate, very possibly ammonium, given that these are blue stars. When I do, I will simply edit this blog entry, so if you are curious about the stars, check back in a few days.

Don't ever let the little kid in you die away. Keep the fascination. Above all, have fun.

Stay green!