KNER propellant is very easy to melt and cast into propellant grains, by far the easiest of any sugar propellant I've worked with. A perfectly suitable propellant can be attained by simply melting a 65/35 mix of KNO3/erythritol at about 225 degrees, stirring once melted and pouring into casting tubes. The resulting propellant will have a density of .059 to .062 lb/ci and be perfectly suitable for use. There are, however, a couple of things that could possibly be improved upon. It would be nice to get the density a little higher and it would be nice to get the bubbles out of the propellant. Of course low density and bubbles go hand in hand, get rid of the bubbles and the density goes up.
KNER density can be improved by leaving the melted propellant at rest in the casting pot at temperature for extended time. This allows the small bubbles in the propellant to rise to the surface and break. Mechanical agitation aids this process, but vibrating molten propellant presents another risk that probably is best avoided. Vacuum degassing really doesn't work with KNER. It would seem under an almost full vacuum KNER boils and foams, at a medium level of vacuum there may be some minimal benefit to vacuum processing. But it is minimal at best.
I know there are people that have used surfactants and/or defoamers in sugar propellants, that always sounded like something I'd like to try, but I really didn't want to have to try ordering small quantities of chemicals from companies that don't like to deal with non-corporate entities. At least initially I'd like to try to find an agent that is available locally and over the counter.
For initial tests I cut a 2" length from some 1.5" aluminum tube, I'll melt and cast into the tube for my samples. First I'd need a control sample that was cast using standard techniques.
Test 1: 65/35 KNER 100 grams Melted at 250-275 degrees, quick melt with no rest time.
Results: 111.6 grams total weight - 32.7 grams tube weight = 78.9 grams or .1739 pounds. Density .06169 lb/ci
Notes: No shrinkage at top, small bubbles throughout with more at top of slug.
Test 2: 65/35 KNER 100 grams melted at 275 degrees, 2-3 drops of Dawn Ultra added after melted.
Results: 112.0 grams - 32.7 gram tube weight = 79.3 grams or .1748 pounds. Density .06201
Notes: Top shrunk in considerably, leading to a crevice that extended about 1/2 of the slug length. Oddly enough even with the recessed top and crevice, the density was higher than the first test. There were however, no bubbles seen anywhere. The viscosity was also noticeably lower, it seems the detergent worked, but what is happening now is the propellant is contracting as it cools. My guess is the propellant always does this, but with small bubbles throughout the propellant, those bubbles give the propellant room or space to contract into.
Test 3: 65/35 KNER 100 grams melted at 250 degrees, 2-3 drops of Dawn Ultra added after melted, quick 3 minute melt, 2 minute rest.
Results: 109.9 grams - 32.7 gram tube = 77.2 grams or .1702 pounds. Density .06037
Notes: Top shrunk considerably with crevice again about 1/2 depth of slug. There may have been a few small bubbles. Probably getting better density at slightly higher and longer melting temps and times. Hard to really say though when the top contracts and a crevice forms.
Test 4: 65/35 KNER 100 grams melted at 275 degrees, 2-3 drops Dawn Ultra added after quick 3 minute melt, 2 minute rest. Slug casting tube preheated to 250 degrees.
Results: 110.3 grams - 32.7 gram tube = 78.6 grams or .1733 pounds. Density .06147
Notes: Noticeably lower viscosity than last test. Same top shrinking and crevice formation as previous tests 2 and 3. No bubbles. I wondered if preheating the casting tube would help by slowing the cooling process, it didn't seem to help at all.
Test 5: 65/35 KNER 1,000 grams melted at 275 degrees, 5.1 grams of Dawn Ultra added after melted. 30 minute total melt time.
Results: 3" diameter x 1.5" core grain cast. Density .062 lb/ci
Notes: It was time to melt a full batch with the Dawn Ultra added as a surfactant, so I cast this into a useable 3" grain. After stirring in the Dawn, viscosity was reduced considerably, but the surface of the melted propellant became almost completely covered in fine bubbles, almost like foam. My guess is that all those bubbles that would have been trapped in the propellant now made their way to the surface. When the grain cooled the top surface again contracted, creating a deep V about 1.25" deep. I had expected that and cast the grain longer than needed and cut the top 1.4" off. Density was okay, but nothing exceptional. Another effect of the contraction was as the grain cooled, it pulled away from the coring rod, creating a core of .05 - .06" larger than the coring rod. Usually the core is only about .02" larger than the coring rod.
Here's a picture of the grain during the cooling process. You can really see how much it contracts as well as the bubbles at the surface.
The Dawn is obviously doing its job of releasing entrained bubbles from the propellant. But now I'm left with bubbles on the surface (now I need an anti-foam agent) and a contraction problem. I considered pouring a grain in stages, allowing time for the propellant to cool and contract before pouring the next stage, I think that would eliminate most of the contraction problem, but with all the bubbles on the surface, I'd end up having a grain with layers of dense propellant interspersed with layers of bubbles.
Test 6: 65/35 KNER 1,000 grams melted at 250 degrees, 4 grams of Dawn Ultra added after melted. About 90 minutes of rest time after melting.
Results:3" diameter x 1.5" core grain cast. Density .06415 lb/ci
Notes: I wondered if melting at a lower temperature would reduce the contraction problem, so this batch was melted at 250 degrees for a prolonged period. Once again I had the problem of bubbles on the surface, so out of curiosity I tried the vacuum pump. It boiled and foamed up more than regular KNER without the detergent, so after trying the vacuum I let the propellant rest again for 30 minutes or so to allow the bubbles to break on the surface. I also tried using some liquid silicone spray to break the bubbles at the surface. The silicone may have had a minimal positive effect but certainly isn't the answer by itself. After casting the grain the propellant cooled and again left about a 1" deep V from contraction at the top, it also again pulled away from the coring rod leaving a core diameter of 1.55". After cutting the V portion of the grain off, the measured density was quite good at .06415 lb/ci. I think the main reason for the increased density is the longer rest time, I don't think the vacuum helped, but I may need to do a few more tests to confirm the results.
Test 7: KNER 65/35 900 grams melted at 275 degrees, 5 grams Ultra Dawn stirred in after 45 minute melt. 15 minute rest then propellant cast.
Results: Density .06266 lb/ci
Notes:There was considerable contraction, as expected. I should have made a 1,000 gram batch since quite a little needs to be cut off the top to get a full grain of good density. This grain was a little short so I still had some bubbles in the top part of the grain. Density was good with no other voids or bubbles detected.
Results: Density: .06366
Notes: I know this propellant will contract a lot, and leave bubbles on the surface. I also know I won't get as good of density out of a small slug as I would a full propellant grain that is trimmed to length after casting. So I added propellant to the slug as it cooled and contracted. The last addition was 12 minutes after the grain was cast. Despite the fact that there was bubbles in the top 1/4" of the slug, the density was still good at over .063 lb/ci.
Results: 108.9 grams - 32.7 gram tube = 76.2 grams or .1680 pounds. Density .05959
Notes: 5 drops mixed in had no noticeable effect on viscosity. Another bad sign was the fact the when I filled the slug tube, the propellant rounded up on top, indicating there was still good surface tension in the mixture. So I wasn't surprised at the low density and small bubbles throughout the propellant once cured. This pretty much acted like normal KNER propellant without any surfactant.
Results: 78.3 grams or .1726 lb. Density .06123 lb/ci
Notes: A lot of steam when the Rain X was stirred in. Density was slightly better with noticeably fewer entrained bubbles. No contraction, but an interesting note is on all these tests with fairly low density, the propellant shrinks away from the aluminum casting tube slightly, allowing the propellant slug to be removed. All the higher density tests had contraction from the top and core. It's very minimal contraction though, as the grain outside diameter is within .01" of the tube diameter. Again, I'm sure it's the entrained air in the propellant that allows the propellant enough room to slide out of the tube. It would seem the amount of PDMS in Rain X isn't enough to do the job, I'll move on to other possible candidates.
Test 1: 65/35 KNER 100 grams melted at 250 degrees.
Results: 74.78 grams or .1649 lb. Density .05849
Notes: This was just a quick attempt at an over the counter anti-gas product. It didn't work, not only did it not reduce bubbles, it also made the propellant much more viscous.
Test Series C: PDMS commercial anti foamer
Results: 77.0 grams/.06053 lb/ci
Notes: I probably melted this at too low of a temperature as the propellant was fairly viscous when melted. No small bubbles but there were larger bubbles/voids that were probably due to the low melting temperature. Virtually no contraction.
Test 2: 65/35 KNER 100 grams melted at 275 degrees, 1.25 ml PDMS added after melted
Results: 77.9 grams/.06121 lb/ci
Notes: Much less viscous at the higher melt temp, but still may be slightly more viscous than plain KNER. Top of grain was somewhat porous but that may be due to the propellant cooling while I tried to scoop the last of the melted propellant into the top of the tube, no other bubbles or voids noted. Very slight contraction. Other than the very top of the grain this slug looked excellent. I think I'll melt slightly larger batches so I don't have to work so hard at getting the top filled.
Test 3: 65/35 KNER 150 grams melted at 275 degrees, 2-3 drops Dawn Ultra followed by 1.25 ml PDMS stirred in after melted.
Results: .06225 lb/ci
Notes: My aluminum casting tube is .01" shorter from sanding the propellant smooth to the top after each pour, so I'll just note ending density from here on. Also note I'm making 150 gram batches now so I don't have to scrape the bottom of the melting pot to cast each slug.
To try to keep contraction limited, I kept the melting pot heated for about 18 minutes after the slug was cast, that allowed me to "top off" the slug as it cooled and contracted. The most contraction took place in the first few minutes, I added propellant twice early on, then again after about 10 minutes and finally at 18 post cast. This slug looked excellent throughout, no small bubbles other than a few at the very top.
P.S. After looking at the slug from this test several days later, the propellant had an unusual molted appearance. Not sure what's going on, but it would seem the PDMS and Ultra Dawn together have reacted in some way. I don't think I was going to pursue this mixture regardless, but now I'm sure I won't.
Test 4: 65/35 KNER 150 grams melted at 275 degrees, 1.25 ml PDMS stirred in after melted.
Results: .06171 lb/ci
Notes: Virtually no contraction. Some bubbles at the very top 1/16" but the rest of the slug had no bubbles or voids at all. Didn't need to top off since there was no contraction.
A picture of the propellant slugs from tests C-1 through C-4. Note sample 3, I broke off the top of the grain trying to pull it out of the casting tube.
This is all very interesting, but I haven't found a "silver bullet" yet. The Dawn Ultra shows a lot of promise. I think if I can get rid of the surface bubbles I'd see much better densities. As it stands now, the combination of Dawn Ultra and PDMS is giving me pretty good densities. About as good as I get with long melt times, followed by vacuum degassing and vibratory table desgassing. What is really intriguing is that I see a lot of contraction with the Dawn Ultra and virtually non with the PDMS, with the Dawn I have surface bubbles, with the PDMS not so many bubbles but still not great density.
It seems to me that the surfactants in the Dawn must be having an effect on how the molecules interact as the propellant cools. Somehow the Dawn must allow the molecules to pack in tighter. The PDMS propellant appears to be bubble free, but has lower density. Perhaps we're talking about entrained gasses at the molecular level. Not my area of expertise I'm afraid so I'll just have to keep plugging away at it the old fashioned way, trial and error!
Series D: Finish Jet Dry
Test 1:150 grams 65/35 KNER melted at 275 degrees, 1.25 ml Finish Jet Dry stirred in after melted.
Results: .06072 lb/ci
Notes: I had hoped the Jet Dry would have some of surfactant abilities of Dawn Ultra without the bubbles. Jet Dry uses non ionic surfactants and is really just a wetting agent. There was no contraction as it cooled, so that was actually an indication to me the density wouldn't be good, and it wasn't. It did seem to have one good effect, it seemed to keep the propellant mixed better. KNER has a tendency to have the KNO3 settle out somewhat and the tests using Dawn Ultra thin the propellant out so this effect is increased.
I think I'm done for the moment. The only additive I found that really increased density and lowered viscosity was the Dawn Ultra. The lowered viscosity does come at a price though. The melted erythritol becomes so thin the potassium nitrate quite readily settles out of the liquid. It then becomes imperative to stir well before casting a grain to make sure you have a homogeneous mixture. The contraction is both good and bad, contraction obviously increased density, but it leaves a void at the top of the grain. So there probably will be somewhat more propellant waste as the top of the grain needs to be cut off and discarded. The good news is that the top of the grain is also where all the little gas bubbles that are released from the propellant end up, so those get cut off as well.