Fuel performance data gathering and grain inconsistency problems.

[Oinikis]

Hello. So I've made an electronic dynamometer and a reusable adjustable rocket motor for fuel performance data gathering, (when having a flush nozzle you can derive from Bernoulli's principle that thrust=chamber pressure*throat area). So using this data I will be able to get some important data like Kn/pressure graph and such.The motor can have five different nozzle configurations, with throat diameters of 4,5 mm; 5 mm; 6 mm; 7 mm; and 9 mm. The motor consists of:

 

40 mm OD, 36 mm ID, 50 mm aluminium tube;

two 50 by 50 mm steel plates, with holes in each corner, and one with a hole in the middle as a nozzle;

4 threaded rods keeping the assembly together.

I have 5 plates for nozzle for changing the throat diameter, and I use 4 layers of aluminium tape stuck to each other as gaskets.

 

For the grain I use BATES with dimensions:

OD of 34 mm

ID of 10 mm

length of 50 mm

Outer surface inhibited.

This configuration gives me surface area of:

Initial: 3230 mm

Max: 3700 mm

Final: 2780 mm

 

With the throat diameters I have that gives us the peak/average Kn's of about 50; 100; 150; 200 and 250

 

My plan is to measure the thrust, calculate the chamber pressure, and assign peak value to peak Kn, because surface area progression profile is quite flat near the peak.

 

Sound's good? not so fast. Of course I got motor working and all, it's working no problem. I did 2 tests on the rig, one with 9 mm throat (Kn of 50) and 7 mm (Kn of 100), and first of all, on the second test the thrust and total impulse was significantly smaller, by like 20-25%, while the burn time and thrust curve shape were similar. With bigger Kn they should have been greater or at least equal. With this sort of inconsistency I can't gather the data. So, here is what I do, and where the problems might lie.

My casting tool is a plank with a spindle sticking out. Then I put the inhibitor inside a PVC tube, and onto the spindle. On the plank there is a widened tube glued in to center it all. I just take the rcandy, ant poke it with a spindle to get it around, and press it down with a stick. This way I get no cracks and air bubbles, and the grains seem consistent. I forgot to weight the ones I tested, but I weighted the remaining 3 (I had 5 for 5 configurations) and their mass was over 70 grams, plus or minus 2 grams.

 

My method of making rcandy is having 65/25/10/30 KNO3/Sugar/Glucose/Water. I put the nitrate and water into the pan, and heat until it all dissolves. Then I add the rest of the ingredients. This way I don't get to boil too much water, but dissolve all the nitrate. The glucose helped make it better to work with, but it would be better.

 

After casting I insert Ematch onto them, send the wires down, and seal them completelly using tape, and store them in a freezer. I might make a desicator for storage later. Anyway, if humidity was the cause, it would have burned for longer, and total impulse woudn't differ that much.

 

Then also the shape of the curves suggest that the grain doesn't ignite uniformly, it seems that core starts burning like half a sec before other surfaces.

 

My main theory is that while I make the fuel, I have to constantly reheat it, and that it might change due to that, although I didn't notice any colour changes.

http://i.imgur.com/PdNK5gPh.jpg

Here is the test stand. Thing above is phone holder for filming.

 

http://i.imgur.com/tOKJuPPh.jpg

 

Here is the motor with all the accessories.

 

http://i.imgur.com/zPoIJj8.jpg

 

And here are the graphs in chronological order. On top you can see the impulses. So what do you guys think? Do you have any questions and suggestions?

Edited December 20, 2015 by Oinikis

[BlueComet24]

What do you use as an inhibitor?

I would like to see you test different fuels, such as sorbitol, xylitol, or erythritol; and different additives, such as Fe2O3, charcoal, or sulfur. I switched to using xylitol and sorbitol (separately, not in the same comp) for rocket candy and now I'm spoiled - It's really nice to work with. I've not actually tested either in a motor, because I've just made small BP motors and used RC for smokes so far, but I've heard it works well. The catch, aside from being quite pricey compared to sucrose, is that sugar alcohols have lower Isps. I have tried using a comp of 65 KNO3, 17.5 sorbitol, and 17.5 sucrose and it definitely was more energetic than 65/35 KNO3/sorbitol and still would melt together, but it was quite hygroscopic.

[Oinikis]

As an inhibitor I use thick paper, but as I mentioned after casting I seal it in tape. Forgot to mention, the inhibitor seems to hold out nice, because every time I disassemble the motor, I find the inhibitor more or less intact, only with ends burned, due to the nature of the grain.

I myself would love to do tons of testing with various fuels and such, but as you see I have some slight problems. But also, making fuel and grains takes some time and resources.

 

http://i.imgur.com/Mgj5VyY.jpg

 

Here is the picture of the chamber after the testing. You can clearly see the inhibitor stood nice. But something else caught my eye. The inhibitor is deeply charred only like for 6-8mm, while the web thickness is 12mm, meaning the core burned for quite some time, before the other surfaces ignited, something you can notice from the graph. So, I need something to ensure better ignition. Maybe I need to have a more powerfull e-match, so it gives a sharp rise in pressure, so hot gasses quickly go everywhere, and ignite the grain more uniformly.

 

Also, I attached a test video in the original post.

[stix]

Good to see you doing some testing Oinikis. I like the re-usable setup.

 

Firstly, are your grains "true" bates grains, ie. the only inhibited surface is the outside of the cylinder, leaving the core and both ends exposed? I'm assuming this is the case.

 

I agree that the 7mm throat nozzle should have increased the thrust and reduced the burn time. It's almost as though the graphs have been swapped around. Are you SURE you haven't mixed something up?. To me it looks like graph 1 is the 7mm and graph 2 is the 9mm. I can't see how you would get less with the 7mm because I'm sure that your fuel making is generally consistent, so it's got me stumped.

 

Secondly, I come up with different theoretical Kn ratios. I wrote some software a while back that does the calculations but haven't used it for a while. Also my motors aren't true bates in that both ends are inhibited giving a progressive profile. Nevertheless I did include in the software to be able to expose one or both ends and I'm sure I checked it thoroughly at the time and it calculated correct.

 

I've attached an image showing the data that I came up with based on your specs. (please disregard the grey column). The first yellow column is the 7mm diam. nozzle and the second the 9mm. I'm not really wanting to double check my calculations again so it would be good if someone else has another way of checking.

Cheers.

 

[EDIT] Just double checked those figures again manually and they look correct - unless of course I keep making the same stupid error.

Edited January 4, 2016 by stix

[Oinikis]

Alright, let's get to it.

 

Yes, it's true BATES grain, the way you described it.

No, I haven't mixed anything up, the grains really aren't consistent, further test showed that, and I have a possible solution, but more on that later.

 

Your calculations are correct, I messed up a little, I haven't checked the exact values. On Jimmy Yawn's page I have found an excel program he wrote, where you put in dimensions, and it gives you the whole graph of progression. The values were the same as in your calculations. I checked the formulas in the program, and well, as expected everything seem ok.

 

So, I decided to do some tests with 5mm plate, and use it to figure stuff out. So, as you can see in photos of original post, grains are wrapped in masking tape. So I put two layers of aluminium tape on the end which faces the nozzle. So in order for the gasses to escape, pressure has to build up first and break the tape. This made initial rise in pressure sharper, which I think is better for uniform ignition. In the graphs you will be able to see the sharp rise.

 

before we start talking about the further two tests, let's talk about the grain. First of all, yes, they are that inconsistent. First of all, the mass varies, and for that I will weight the amount of fuel I need just before casting. Then another thing might be that the fuel isn't pressed properly. The method of mine is a base with the spindle, tube with the inhibitor around it, and I just poke the fuel around the spindle and press it down with a small dowel (non cored, it fits between the spindle and the tube). That might lead to uneven end surfaces, and not pressing well against the inhibitor. Now that I think of it, what was I thinking?

 

So now I made the tube longer (6 cm, while the grain is 5 cm long) and a short cored dowel which is a good fit. So my plan is to weight out exactly 70 grams of fuel, then put it in like I did before, and then put the cored dowel on top and smack it with a hammer, so ensure the proper shape and consistency. The extra centimeter is so the dowel is in correct place. I try it this way, and see how it goes.

 

Actually, I did three more tests, but first of them was CATO. I think the grain might have fractured, or the inhibitor didn't inhibit enough, cause it burned way too fast. The estimated chamber pressure was Actually I'm kinda glad it happened, because it showed that my rig was prepared for that, I made the new aluminium tube bit longer, and bought some screws, instead of threaded rods, so I could tighten it much better, because these threaded rods suck, the threads are very weak. The estimated pressure on that was about 60 Bars, but with improved setup on further tests I reached over 50 Bar and it held no problem, because I tighten it much better

 

Alright, so first things I noticed during these two further tests, that they had frickin Mach diamonds. Next, there are cardboard pieces lining the inner surfaces of the rocket holder in case of CATO. They were pulled in by the Venturi effect created by the exhaust. I just didn't think the effect would be that strong. Another thing you can notices is the vibrations of the camera and flashing exhaust flame. You can see it very well on the graphs. The pressure had to jump up and down by about 10 Bar to produce such oscilations. I suspect they are formed due to the nozzle being flush, and the throat being so short. I think I won't have this with a proper nozzle. for now I can take the averaged out value for my data gathering.

 

Then, while the video footage of both tests looked similar, one had Isp of 66 seconds, while other had 79 seconds. I checked and checked, but I can't find a mistake, also the graphs look drastically different. This really confuses me. Also, the grains are quite old, but they were kept in the freezer and were sealed with tape, so them being this cold might have an effect, but I doubt it. Also note that these two last tests were done in some serious cold, that is -20 C, but I doubt it it had any effect on the electronics and the motor itself.

 

So, now I will attach all the things.

 

http://i.imgur.com/vlqIgAS.jpg

 

 

 

 

http://i.imgur.com/jbl3bgd.jpg

Edited January 6, 2016 by Oinikis

[stix]

It's always going to be difficult when the grains aren't consistent. We need that, otherwise we cannot proceed to a point of meaningful evaluation.

 

Personally I would NOT put the grains in the freezer. A simple ziplock bag with some desiccant would be better. There could also be some friction due to your horizontal test apparatus. If you are really committed to it, then use some graphite powder to ensure you have less friction. Otherwise (and IMO better) is to re-configure for a vertical test stand. No friction, and the fuel loss can be mathematically accounted for.

 

Cheers.

[Oinikis]

As for the rig. I was sceptical too at first, but then I measured the friction coefficient and calculated that maximum losses due to friction with some misalignment would be really small, but don't forget that the holder is pressed against the load cell, so it doesn't have to move. I have tested an tested, and the measurements seems to be really consistent.

 

Just recalculated, maximum friction would be someting like no more than a newton, and the thrust we are having is about 100 newtons, so yeah. But still, since it doesn't have to move and is pressed against the friction doesn't seems to really apply.

 

The ultimate solution in my opinion doesn't have to be a vertical rig, but I imagine a rocket holder on a huge arm with a hinge, pressed against the load cell, but I'm still convinced my design is good enough. Although i'm often wrong.

 

 

Now as for the fuel and the grain. I used 65/35 KNSU mix. to get it softer I replaced some of the sugar with glucose, and use 65/10/25. But if you look at the formula of glucose and compare it to sugar, glucose has more oxygen in it, and more hydrogen, which requires less oxygen than carbon, so overall when replacing sugar with glucose, we might have to have bit more glucose. Anyway, i'm not sure, so i'll try 60/20/20 KNO3/sugar/glucose and see whether it is easier to work with, and whether the specific impulse suffers or not.

 

 

Also, to get the graph, I might only need well two Kn's something like 100 and 200, because the graphs is almost linear, a slight curve and crosses point 0;0, so only with 2 point we can have a good estimation of the curve, and use other Kn's for checking. I think if we focus on only these two points, we can get more accurate results by doing more tests, and get more accurate graph. Well, first of I'll try to work out the grain, and then I'll do the testing.

[Oinikis]

Okay so I tried the new things, and I failed epicly, I was so upset, I didn't even curse. I'm tired of all that cooking and casting, it is hard, and I can't get it really consistent, thus, I have yet another crazy idea.

 

Okay, first of, screw glucose. Mix some 65/35 KNSU, throw it into ballmill, add a very small amount of water, like 1-2%, just so the powder has no air, and has fixed density, put it into molds, weight each mold before and after, to have the right amount, and put them into oven. After some time water will evaporate and the sugar will start to melt, that's when you take them out, let them cool, and extract. if this works, it will be a superior way to make the grains, it requires much less work, isn't stressful, the grains would be really consistent, and you could make much more at a time.

 

Of course, one wouldn't use the same oven as your mom/wife/creature uses for baking cookies. You could use these mini ovens used for making sandwiches. I saw James Yawn uses one for reheating his Rcandy. But, the old electric stove I use has an oven in it, just the electronics is screwed up. So I could strap an arduino to it, and make a temperature regulator.

 

So what do you guys think about this idea?

[stix]

Okay so I tried the new things, and I failed epicly, I was so upset, I didn't even curse. I'm tired of all that cooking and casting, it is hard, and I can't get it really consistent, thus, I have yet another crazy idea.

 

Okay, first of, screw glucose. Mix some 65/35 KNSU, throw it into ballmill, add a very small amount of water, like 1-2%, just so the powder has no air, and has fixed density, put it into molds, weight each mold before and after, to have the right amount, and put them into oven. After some time water will evaporate and the sugar will start to melt, that's when you take them out, let them cool, and extract. if this works, it will be a superior way to make the grains, it requires much less work, isn't stressful, the grains would be really consistent, and you could make much more at a time.

 

Of course, one wouldn't use the same oven as your mom/wife/creature uses for baking cookies. You could use these mini ovens used for making sandwiches. I saw James Yawn uses one for reheating his Rcandy. But, the old electric stove I use has an oven in it, just the electronics is screwed up. So I could strap an arduino to it, and make a temperature regulator.

 

So what do you guys think about this idea?

 

I feel your pain brother. It's always difficult to maintain consistency.

 

Your idea of putting powder into the mold, then into an oven is similar to something I tried a few years back. I found the main problem is that during melting it tends to separate leaving some liquid oozing out (unless you can stir it during the melt). Another issue is that you will need a much longer mold because the powder will become dense when melted and drop in level in the mold. Also all your tooling will have to be able to handle 186C. The idea has merit but unfortunately as usual, it's not as easy as it seems.

 

Alternatively, are you able to get some glucose powder? You should be able to get it at the supermarket or health food shop. I've found that it's around five time the cost of table sugar - unless you are going to make "huge" rockets , then for most of the rockets we are making, the extra cost has it's benefits.

 

The main benefit being that it melts around 146C. What this means is that you have more time to mold before it "sets up". I've found that I was able to make a "putty" that I could handle with my bear hands (although still very hot) then make "snakes" and push into the mold. It's worked out well. Once I've done this I put the whole apparatus into the oven to slightly soften it (125-130C) remove and hammer press using a donut shaped dowel. This way there are little to no voids.

 

My current fuel (untested in a motor yet) is KNSUSB. KNO3/Sugar/Sobitol 62/20/18. This fuel can be melted and mixed at 95C. I use icing/confectioners sugar. Sugar does NOT melt at 95C so therefore I have a fine powder mix of KNO3 and sugar "suspended" in the sorbitol. This remains very soft and pliable and able to be molded and handled easily. It sets up to some extent within 5mins. 24hrs later you can still bend a strand of it if you carefully bend it. I like it!!!

 

It's like a never-ending battle, frustrating at times and very rewarding when things do go right. I reckon that from what you have said, your test rig is working well enough. When your test rig is working and you can evaluate the data, then you ARE seeing the inconsistency of your fuel and/or methods. Which is sort of why you made the test rig in the first place .

 

I'm sure you'll get there.

 

Cheers.

Edited January 9, 2016 by stix

[Oinikis]

Sorbitol seems, nice, i'm happy for you that you found something that works.

 

But as for that method. You missed one point. Before putting the powder in, you wet it with water, like 1-2%, that way, it collapses and air get removed from it, and it achieves that density. This way it won't shrink, and be uniform, at least I hope so. Some time ago, due to unknown reasons, I made BM using KNSU instead of BP. It required very little water, powder collapsed when water was applied and formed a thick dense goo. It dried to a hard, white solid, which didn't shrink (Well, of course it did shrink, but only due to loss of water, which there was littleof to begin with) nor change shape during drying, and burned in a fashion similar to rcandy. The fuse I made also was quite good, I used to wrap it in tissue and use as a fuse. Well, I think I'll make a small amount and just try it in a tube, and see how it goes.

[stix]

Sorbitol seems, nice, i'm happy for you that you found something that works.

 

But as for that method. You missed one point. Before putting the powder in, you wet it with water, like 1-2%, that way, it collapses and air get removed from it, and it achieves that density. This way it won't shrink, and be uniform, at least I hope so. Some time ago, due to unknown reasons, I made BM using KNSU instead of BP. It required very little water, powder collapsed when water was applied and formed a thick dense goo. It dried to a hard, white solid, which didn't shrink (Well, of course it did shrink, but only due to loss of water, which there was littleof to begin with) nor change shape during drying, and burned in a fashion similar to rcandy. The fuse I made also was quite good, I used to wrap it in tissue and use as a fuse. Well, I think I'll make a small amount and just try it in a tube, and see how it goes.

 

I don't think I missed any points. I shouldn't have mentioned sorbitol. Who cares? sorbitol, shmorbitol?

 

Just make powder and smash it down with a hammer, then drill a hole - works for most people. Beyond that, then we experiment because we want something better.

 

Best to go along your own path