11/2/01

The test of 10/27 was successful enough to warrant further experimentation.  I wish to find a more elegant way to ignite these grains without resorting to massive charges of black powder.

But first,  a new batch of candy.  I tried pouring this large batch into three rectangular Pyrex pans rather than two, suspecting that my difficulty in getting a fast-burning fuel in large batches stems from having the liquid too deep in the vessel, thus slower to dry.

Well, it worked!  This batch is a nice light color and burns at 11 seconds per inch uncontained, which is the fastest I've obtained from a batch this size.  I did rotate the pans from the top to bottom shelf, as the one on the lower shelf cooked faster.  The yield is about 650 grams.  Now if I buy one more flat pan, I can make 2 pounds at a time.

This is made according to the process described on the Rcandy page.

So now for the lacquer experiment.

Here I plan to make some F-class uninhibited grains, dip them in nitrocellulose lacquer, and fire them in pvc casings.

Just in case anyone is interested, here is how the grains are formed.  I have cut some pieces of schedule 40 pvc, 1 inch diameter and 4 inches long.  Each is lined with two strips of typing paper cut the short way, rolled up and inserted into the case.  This is to create a slightly undersized grain that can be coated with lacquer and still fit into a 1 inch tube.

I want a bit of priming embedded in the head-end of each grain, so I choose nine nice chunks of homebrew black powder and place three in each tube.  Note that I use a stick to distribute them toward the outer edges of the tube.

Now the candy comes out of the oven.  It has been heating at 200 degrees for awhile, and has a nice taffy-like consistency.  I had previously rolled it into balls of about 80 grams each, knowing that each tube would take about that amount.  A piece is pulled off and dropped into a tube, where it is pressed firmly with a dowel to make sure that it envelops the top side of the black powder grains.

More candy is dropped into the tube until it is almost full.  Then a 1/4 inch wooden dowel is rammed through to form the core.  the candy is packed down with another dowel, attempting to get air spaces out.  I really need to work on this little bit, as you will see in a moment.  The grains are allowed to sit in their tubes about 1/2 hour, at which point they are warm but firm.

In the meantime, I must make some ignitors.  Here I am using copper filaments from a section of old lamp cord.  Nichrome is not required, as I plan to use house current. This makes a bright, green spark.  I clean each filament by drawing it through a folded piece of 600 grit wet-or-dry paper until it is shiny and wrap it around the short lead.

Then it is soldered,  wrapped around the other lead until it reaches the stripped end, and soldered there.  I like solder.  Solder is good.  Those who simply twist get what they deserve - flaky connections.

Please help me celebrate this very special day:  My $2. Shadio Rack soldering iron got its first new tip after 27 years of service.  The old one was down to a nub, and was retired with honors.

How's this for a shunt - soldering the leads together!  Just the tips, though, so they can be pulled apart when ignition-time comes.

The ignitor pictured above is a bit long for my current project, so I made a few shorter ones.

A grain is removed from its tube, and the paper peeled off.  Well, about half of it.  Next time I will use something less porous - do I correctly recall that someone on Arocket mentioned casting in Mylar sheets?  I need some!  The ignitor is run up through the grain and bent over one of the black powder warts.  A small piece of hot candy is pressed on top, embedding the ignitor in the grain and sealing it in place.  I want to be sure this spot gets lit!

These are possibly the ugliest grains I have ever made.  My deepest apologies to any gentle viewers for subjecting you to this ordeal.  And a few more apologies for the out-of-focus pictures - it's kinda hard to take photos while doing other stuff, like driving, or making propellant grains.

But I see no reason why these grains would not work - in my low-tech engines an extra fold or two should not make a lot of difference, so I will go ahead and use them.  More significantly, I am hoping that the residual paper does not inhibit ignition too much, as it still covers about half of each grain.

Message from the future:  The sticking paper problem was easily solved.  I took some 2-inch wide clear Scotch mailing tape, covered a piece of paper with it, cut it into 4" strips, and used this for the liner.  It peeled off the cooled grains with no trouble at all.

Since the grains were a bit warm when I took them out of the tube, they have distorted just a bit.  Strange that recrystallized KN/Sucrose takes awhile to "set" perhaps like KN/Sorbitol but in minutes rather than hours.  Upon first cooling, it is somewhat flexible, but after sitting an hour or so it becomes quite hard.

Nice day, so I go outside to do the smelly stuff.  Painting the first grain with a brush was labor-intensive, so I have poured an olive jar full of NC lacquer so that I can just dip the grains.  Ignitor leads make a lovely handle, and then form into a perfect hanger.  Anyone who didn't know me might think I had planned it all out.

First, a quick check to make sure the grain will still fit in a 1 inch tube.  This one does.

Now the first dip.

Each gets dipped, hung out to dry for 1/2 hour, then dipped again.

Too bad Halloween is over.  These grains would have made a nice decoration.  Just as well.  No kids ever come here to trick/treat - this place is really scary.  I had to eat all the chocolate myself.

Note that the lacquer has skinned-over the core at the head-end.  I poked this out, as I want full circulation through the grain, both for drying, and for ignition.

Aquatic Test

While the lacquer dries I must find a way to entertain myself.  I wonder if I can ignite lacquered candy under water.  Some hot candy is wrapped around an ignitor with a pinch of black powder.  The candy is dipped in lacquer and allowed to dry for a few minutes, placed in a plastic bottle of water, and the switch thrown.

There is a snap, the bottle turns over, and the water runs out.  The candy did not ignite.  Examination reveals that it had split open, and was broken into three large pieces.  I have two theories, and have serious doubts about both of them.  One is that the candy never ignited - the pop of the black powder was so vigorous that it did not have time to catch.  The other theory is that the candy did ignite, but the onrush of water put it out.

Anyway, I set fire to that piece and despite being wet it burned just fine.  Waste not and all that.

So now for the main event.  I have constructed a case of 1-inch pvc, 5-1/2 inches long and with a pressed clay nozzle.  Throat diameter is 5/16 inch.  The fuel charge weighs 75.0 grams including the ignitor leads.  Ignitor weighs-in at 3 grams, leaving 72 grams for candy, priming, and lacquer.  I wish I had weighed the grain before dipping - it would be nice to know how much nitrocellulose is there.

Click on the photo below to download a video of the burn.
(Beware:  600k download for 3 seconds of video)

The "click" heard at the beginning is the "ignition switch" (actually the "reset" button on a power strip.*)  There is a puff of smoke at 0.2 seconds, then the burn begins at 0.9 seconds.  The sound rises to maximum almost instantly, and continues for 0.7 seconds.

(*The switch-click has proven very useful.  I think that when I finally build an igniton controller, it will include an auditory signal to make it clear at which instant the power was sent on its way.)

My take on this burn:

I assume that the first puff of smoke is from the black powder under the ignitor.  Apparently this little bit of priming is not adequate to ignite the nitrocellulose lacquer, but the candy has been lit at the BP spot(s).  The flame spreads, building pressure and venting toward the nozzle.  At 0.9 seconds either the lacquer ignites setting fire to the fuel under it, or the candy reaches critical pressure and burns away the NC lacquer.

Given the suddenness of ignition, I suspect that the nitrocellulose lacquer might actually facilitate ignition.  Further tests are planned, using a thicker coating.

The ignitor wires are stressed-out, missing insulation in several places.  I am not sure whether this is from being anchored to the grain as it began to burn, or if the solvents in the nitrocellulose lacquer dissolve the insulation.

(Update from the future:  I examined some leads which had been dipped but not exposed to flame - the nitrocellulose peels off easily, the insulation is not damaged or degraded as far as I can tell.)

Further tests:  Is nitrocellulose lacquer a good moisture barrier?

It occurred to me that if NC lacquer is impervious to moisture, I should be able to place a coated grain in a glass of water overnight and it will be OK in the morning.  So I made two little grains, just slugs of candy with an ignitor and a wart of BP inside, double-dipped in NC lacquer.  After drying for a couple of hours, they are dunked in a big mug of tap water.

After a couple of hours in the water, both grains seem intact.  Encouraged, I decide to put one of my "F" grains in too.

Well.  During the night the candy in the F grain got water in it.  It ran out like a firehose when I picked it up.  I suspect that the ignition lead broke the lacquer seal as it was handled rather roughly - I won't say by who!.  Fortunately, I make recrystallized candy, which can be reincarnated.  The water was strained through a coffee-filter to get out the greeblies, baked in the oven to dry it out, and kneaded with a spoon.  This was used to make another grain on 11/4, which worked just fine.

Next morning, one of the aquatic slugs feels a bit squishy.  The other is hard as a rock.  Apparently, there was a small opening in the NC lacquer at the head end of the one on the right.  The ignitor popped it open but it did not burn.  The lower one burned really well.

This proves two things to me:  NC lacquer can be a really good moisture barrier.  And it can be really bad if it has any kind of hole in it.

More experiments 11/3/01

My biggest wonder about the previous tests is whether the lacquer is inhibiting ignition, assisting ignition, some of both, or something else entirely.

So I make another grain just like the ones from yesterday, but with two differences.  First, it is not coated with anything, no lacquer, and none of my customary black powder priming on the outside.  And it has only one large chunk of black powder embedded in the head end rather than three.
 

Click on the photo below to download a video of this burn.
(323k download for 3 seconds of video)

In this test, ignition was about as quick as I have ever seen.  Analysis of the video sound file shows a delay of only .14 seconds between the click of the switch and full thrust.  The thrust-burn lasted 0.66 seconds.

But a similar test on 11/4 had very slow ignition, and a third one very quick  The difference?  The slow one had an incompletely-embedded ignitor - a small crack exposed the black powder grain and the tip of the lead wire.  I suspect that this allowed gas to leak gradually rather than building up and releasing suddenly as the candy "cap" fractures.

Results up to this point:

1.  Nitrocellulose lacquer seems to be a good moisture barrier for candy grains, provided that it is a complete coating uncompromised by cracks, pinholes, or other leaks.  A less-brittle coating would be desirable, as this lacquer seems to crack easily.

2.  NC lacquer seems to inhibit candy ignition, but only a little.  Full ignition was achieved in all tests, but after a short delay.

3.  Embedding an electric ignitor just under the surface of a candy grain seems to offer very quick and full ignition.

These results are very incomplete, and mostly suggestive of possibilities.  I welcome comments or suggestions of any kind on this work.

James Yawn
jyawn@sfcc.net