Saturday, 21 January 2012

Haven't been up to a whole lot lately. I ordered all the parts I need to build the peroxide test stand last monday and am waiting on them to arrive. Hopefully they will be here before the end if the week so I can start work on the test stand that weekend. I decided to reuse the thrust chamber from the nitrous-ethanol engine which should speed up construction. The control system still needs a little work, particularly in the reliability of the communication protocol. The strings over serial work quite well, but I would like to add in some reduncencey. The electronics system also needs some work but more on that later.

 Last weekend I also experimented with other methods of evaporating peroxide. I tried plain evaporation at 75 degrees and forced evaporation (blowing a fan on the liquid) and was quite surprised with how well even plain evaporation worked. In about 10 hours about 2L of peroxide went from 45 to about 75% with just plain evaporation, with a stainless bowl about 45cm in diameter. I stopped forced evaporation after dust and bugs found their way into the peroxide, and started to fizzle. The hydrometer I have only goes from 1.2-1.3 Kg/L so I am not sure of the exact concentration but it was just above 1.3. I had previously thought that the losses from decomposition would be too great (even at low concentrations) to make the method viable, and it still might be at higher concentrations but it seems like it will be an efficient way to get to %80 which is quite usable. I also bought another 25L container of %50 last weekend as I was running low. I really need to do some experiments varying temperature and surface area to try to work out a relationship between time, concentration and yield. I also came across an interesting paper which uses a desiccant suspended just above the surface of a peroxide solution to concentrate it. The paper suggested that it worked quite well, although was slow at room temperature. Even if it did take a few weeks to concentrate, I could just leave a large amount concentrating, changing the desiccant when necessary.

I seem to have found a source of %90 commercially which I understand is propellant grade. There are still a few unknowns but If all goes to plan I will be buying 40L early next month. Including shipping it works out to only slightly more expensive than buying %50 and concentrating assuming a %35 yield. I think the yield from DIY would be a bit better than %35, so cheaper not counting the investment I have already made in equipment and the cost of my time, but I would much prefer to buy if I can. Also from what I understand the %90 is propellant grade so I will be able to use a silver catalyst later if I like. Still I would like to not have to rely on a supplier so I will continue to experiment with concentration.

Tuesday, 10 January 2012

lander concept

Buren has done up a basic design/layout of what we would like the lander to look like. It looks quite cool!

Last night I did some investigation on the air system and made some interesting observations. I measured the flow rate more accurately and confirmed that the flow rate was about 1/4 less maintaining 60PSI than at 30PSI, even after taking out the needle valve. I know AC motors should have a constant speed, and Ariel had the good idea of measuring the speed of the compressor (RPM) at low and high pressure. Luckily the compressor shaft was exposed enough to attach a tachometer to and I discovered that, as expected the speed of the compressor was not significantly different at 30 and 60PSI (maybe %1-3 difference). So if the compressor is pumping at the same speed then that means there must be some losses in the system. There are several this could be but Ariel came up with the most likely, which was that the piston ring leaks more at a higher pressure. I am currently running two compressors in parallel, a new one and a quite old one I got second hand. It is probable that the old one leaks more at higher pressure. There are also thermal losses which increase for higher pressure ratios as the air is hotter when its compressed, but then cools down and decreases in volume but these are not significant. Also I discovered when researching the air compressors that I have that they can actually run continuously. I was hesitant to leave my new compressor on incase it overheated so I atached a thermometer to it and found that it stayed at a constant 50 degrees c. This means I have effectively doubled the amount of sparge air I have available.

I also have been designing the plumbing for the peroxide test stand. For the nitrous-ethanol stand I used steel "bundy" tube which was cheap but a huge pain to work with. For peroxide I am going with stainless braided tube with a teflon liner. Since I am buying all the parts from overseas I wanted to make sure I had everything before I ordered them so I did the drawing in detail. It may be difficult for anyone else to read.....

Monday, 9 January 2012

More Air

I have been trying to improve the sparging process, as currently it takes about 2 days of straight sparging to produce %80 peroxide. I have tried a few improvements, but they have not had a huge impact. I had previously been operating at 50 degrees c, which I increased to 60. Increasing the temperature does increase the rate of evaporation quite a bit but it also increases the rate of decomposition. I don't want to go any higher than 60 for safety reasons. The flow rate of air effectively used for sparging is about 75L/Min, and I think that if I want to go any faster I really just need more air. If we assume that the air is mostly dry, it can hold about 55 grams of moisture per meter cubed. If I have a flow rate of 4.5M^3/hour then I am only loosing 250 grams of water/ peroxide per hour, which isnt great. I did have a needle valve setup after my regulator to vary the flow to achieve the compressor duty cycle I need (%50), but this seems to have been a huge restriction on the flow and after I removed it the flow seemed to increase the rate quite a bit. Despite having done two subjects on fluid flow, it still confuses me sometimes. I would have thought that since the compressors operate at a constant speed, they will produce the same flow rate of free air no matter what the upstream pressure is (until the motor stalls). It does however seem that there is more air flow now that I removed the needle valve. I originally put it in because in order for the regulator to be used effectively there needs to be a pressure drop over whatever it is plumbed in to. Indeed when I remove the needle valve the regulator reeds nothing despise being set on a particular pressure. It is possible that it only seems like there is a greater flow rate, as I didn't measure what they duty cycle was after removing the valve. As best as I can understand it:

There is definitely a pressure drop over the system, but all I really care about is free flow rate (as long as there is enough pressure for the dryer to work properly)

The free flow rate should not change because of the pressure drop unless the flow chokes, which will happen eventually but I wouldn't have thought for such a small system.

So unless the valve is choking the valve somehow but I cant see how. Even if it was It shouldn't affect the flow rate as it will just act as a restriction, however the compressors will still operate for the same amount of time.

I think more investigation is needed.

The other thing I am unsure about is if my refrigerated dryer needs a certain flow rate to function properly. I know it is important to size a dryer correctly for inlet conditions (so you have a big enough dryer) to cool all the air but what if you have a dryer which is too big for your flow rate. It would make sense that it should still work as all it is doing is cooling the air to condense water, but I know that some designs actually use the flow of air to trap the water aerosol. I imagine that if the flow through whatever mechanism (maybe a coalescer) isnt fast enough it might not capture the aerosol.

I have been on the look out for a larger oil free compressors to speed up the process. Anything bought new in Australia is out of the question as it is out of my price range. I have been looking at some Chinese oil free Dental/Medical compressors and it will work out at about $1500 for 20CFM. Short of a huge 100CFM compressor no-one seems to make large ish oil-free compressors, instead they just mount many smaller ones in parallel which looks peculiar but I imagine makes sense economically.

They also sell nice regenerative desiccant dryers which I will need to get eventually. The refrigerated dryer I have now produces air with about a -20c dew point and to concentrate to 90 I will need that much lower (would like at least -80) I am looking at a 50CFM dryer which should last me for quite a while which is aobut $1200 inc shipping. For anyone who doesn't know Alibaba is a site which lists lots of asian suppliers. Most of them have high minimum order quantities but some will sell items individually. 


The other option is to leave the sparger going while I am not there. Yes, that sounds dangerous (and it could be). The idea would be to put a vent on the roof so I could sparge inside, have an extra blower to extract fumes and have everything contained inside a large vessel so if anything leaked it would only go into the containment vessel. I would also monitor it remotely via a camera and would be able to shut it down remotely. There would also be a  system which would inject water into the vessel if a fire was detected. Not my preferred option but worth considering.

Saturday, 7 January 2012


I am currently at the workshop working on some stuff while the sparger is going its thing. I have moved the setup outside and now monitor it remotely via a camera. Having it outside does have the advantage that I can only sparge at night when know one else is around.

Now that I have a steady supply of peroxide I ave started thinking about what to work on next. What I would like to work towards is a hovering vehicle similar to the first vehicle armadillo made, with 4 small thrusters  each with a pulsed solenoid for throttle control. I have always been interested in the control side of things, and rather than building something that goes high or fast I would greatly prefer the challenge of building a actively controlled vehicle. People seem to waste so much time and money on rockets that they never recover.

I am thinking of going for a design with three thrusters in a triangle. Although one less thruster would slightly simplify the design (and save weight) it will complicate the control as the thrusts will not be orthogonal to each other. It shouldn't complicate it greatly as I should be able to write a function which I will give cartesian vectors and will return the required throttles. I have done quite a bit of work with omni-drive robots and this should be quite similar, although there is an extra (coupled) degree of freedom. Getting that to work in practise is another matter entirely.

Pretty much the entire vehicle needs to be designed around the tank('s) as there are very few options available, short of getting something custom made. There are plenty of long skinny tanks for hybrids etc, but there are not manny med-high pressure (1000 PSI) tanks around that are short and fat. (I want the vehicle to be as flat as possible to make it difficult to flip over). There are some really nice 1.1L (1.43Kg %90 peroxide) composite tanks designed for paintball but as I will explain later thats not quite enough volume. The next size up is a 10Lb now bottle which is what armadillo used on their first vehicle and is what I will probably go with. Nitrous tanks never seem to list their volume, only how much now they can hold which i suppose makes sense but is annoying. Considering the density of nitrous oxide the tanks seem to be able to hold about 8.5L although I know that you can not completely fill nitrous bottles so it may be more. I am going on the assumption that a 10Lb bottle will be able to hold 10Kg of peroxide although it should be able to hold more. The bottle itself weighs 6.8kg so lets say 7.

We are up to 17Kg with just tank and a fuel load of fuel. Lets say that each thruster (including consumable catalyst) weighs 1Kg and each  solenoid weighs another 1Kg thats 23Kg. Then there is just structure, batteries, electronics, wiring and plumbing. I am hopeing to keep the whole vehicle under 30Kg. Some of the propellant tank will be used for storage so each thruster will need to be about 100N.

If we assume a modest ISP of 80s, with max thrust we will have a fuel flow rate of 0.375Kg/s, with means that at max thrust and assuming only 8Kg of fuel it will be able to fly for 21.3 seconds. I would be happy with 10 seconds of flight so this scale of vehicle seems to be good, and with 8kg of peroxide per flight I should be able to get in a few flights per sparge batch. I don't want to build something that has outstanding performance, my main objectives are price and simplicity.  I am tempted to try a gimbal, but that introduces a lot more complexity. I would like to go for the smallest vehicle feasible but I don't think I will be able to get more than a few sounds out of the paintball tank. Although that ratio of tank mass/fuel is about the same I don't think it will be easy to build a light enough thruster and I am already going with the smallest solenoid available. Buren is doing up a model of a basic vehicle with should give us something to start from.

So I guess the first step is engine. Convientely  the 100N nitrous-ethanol thruster was designed for 100N thrust, so it will make for a good test engine while I wait for parts to repair the lathe. All I need is a small sleeve to provide an internal edge for a catalyst pack to but up against, a piece of aluminium tube would work well. Really what I need to start work on is a test stand.

I have been doing some experiments with different methods of using potassium permanganate as a catalyst. A silver catalyst has its advantages but everyone who uses them seems to spend so much time trying to get them working properly, and it seems to be more art than science. Poisoning is a big issue especially if you get different brands of peroxide. It seems like although it is slightly more work using a consumable catalyst is the way to to. I would like to avoid using a liquid catalyst (injected into the chamber) if I can, because that would mean a whole extra set of plumbing. Stacked discs of something coated in a consumable catalyst seems to be the most popular design for cat packs, but I would like to use a granular catalyst that I can just sandwich between mesh. I haven't seen any other engines use this design, but it seems to have the advantage that you can just pour in the catalyst instead of delicately arranging lots of discs.

Today I tried making some granular catalyst by letting kitty litter absorb a permanganate solution, then evoperating the water leaving a porous granular material with permanganate inside of it. I should add that the idea of using a absorbent material and evaporating the water came from an acquaintance of mine (thanks!).

I tried two different types of kitty litter, a clay based one (Brown and on the left) and a gel based one (clear/Blue and right). The clay wasn't particularly strong to begin with but after baking it became even softer and clumped together when saturated so the final product contained quite a lot of soft power.   The gel one was hard to begin with and was able to absorb more solution giving a product darker in colour which would indicate  more permangenate (or maybe it lust looks like that because it was clear to begin with). The gel one seemed to be more reactive (both were more than good enough to use for a cat pack) when exposed to %55 peroxide. I would also like to try  the same thing with silica gel. If it is as good as the gel kitty litter it would make an excellent catalyst  material because unlike the litter gel it would make a more uniform material with constant pressure drop between batches (the litter is jagged and not uniform). I think the gel kitty litter is actually silica gel, but it is probably a lower grade than that used for dehydration, which means that dehydration gel might be able to absorb more. There is also a paper based kitty litter that I would like to try. I am slightly worried that the catalyst might just absorb moisture but i suppose that if its operating in a engine there won't be any water to absorb.

Tuesday, 3 January 2012


From the comments on the last post there was some concern about the safety of my setup. I really appreciate these and please  keep these coming because criticism is a excellent way of finding a potential problem.

 I plan on using a consumable catalyst

Sparging in a pressure vessel:
The pressure vessel does have a relief valve, currently set to 130PSI. I should probably lower this, I think all I need is a new spring. After some peroxide came out under pressure however I decided is probably wasn't a good idea to sparge in a closed vessel.

What am I doing to the air:
Two oil dental air compressors (oil free) feed into an an air receiver. The air is filtered by a 1micron filter then fed into a refrigerated air dryer. There is also a 5 micron pre filter on each compressor filter. Currently the setup is running at 60PSI which should put the dew point at around -15C. I plan on increasing the system pressure as the concentration increases to get a lower dew point.

I have tested the setup extensively with distilled water to see if I could catch any contaminants collecting, couldn't see any.

This is probably my greatest concern at the moment. After I initially realised the exhaust was an issue I was bubbling the exhaust through water and venting the exhaust outside (that was when the vessel was closed, with one in and one out pipe). With the open vessel I simply put the entire setup outside and had a large fan to blow fumes away from anyone, which seems to be quite effective and we can't smell anything inside the workshop. Also I only sparge at night when there is no one around. I would be interested in finding out how bad the vapour is for you. VHP is used as a disinfectant so it can't be good, I just figured that because the vapour was not particuary concentrated (initially at least) and that it was being diluted with so much air that it wouldn't be a problem. I suppose I was lucky that my peroxide wasn't too concentrated before I realised that the vapour was an issue.

I think what I might do is:
Sparge outside, further away from the workshop.
Turn the sparge gas off when I need to inspect the rig.

I had the understanding that it was best to filter out contaminants using IX when the peroxide is at as low of a concentration as possible. That is because the IX resin can act as a fuel (i think its plastic) and the IX process can generate heat. Is there any other reason why it would need to be filtered after concentration?

It seems like vacuum sparging would also be a better method because it may increase the speed of concentration., but not necessarily from a safety perspective. The problem people who vacuum distill their peroxide have is that if they loose vacuum, the concentrated peroxide vapour that is safe at vacuum will decompose because it is suddenly at a much higher pressure. That probably won't be an issue as there won't be nearly as much vapour as if the peroxide were boiling. I would appreciate any references anyone has found on vacuum sparging.

Monday, 2 January 2012

After a bit of a rocky start the sparing setup is going relatively well. Last night starting at about midnight and going to about 5am we concentrated 10L from about 45 ish percent to 54 percent safely. It seems that the setup can concentrate around 10% in 5 hours (obviously this will decrease as the peroxide increases in concentration). Up until now my concentration has been on getting the setup working safely, but for the next sparge I intend to take readings of liquid level and peroxide concentration regulary. Eventually I would like to be able to calculate how long a given volume of peroxide will take to concentrate.

I have settled on a completely modular setup using 19L beer kegs. The kegs have two quick connect fittings and one large hole used for cleaning. The brilliant thing about kegs is that they have a stainless syphon going to the bottom which can be used both for bubbling air through and for getting the concentrated peroxide out for use.  To get liquid out you simply need to pressurise the keg gently through the kegs gas fitting. Its good not to have to worry about contamination from adding a sparge line and other bits to the vessel and when the sparge is complete all that needs to be done is close the keg up and put it in the fridge. The kegs also make excellent storage containers.

Initially I had intended to keep the large hole in the top of the keg closed and pipe the has out of the keg through the gas fitting. This enabled the gas to be easily piped outside and also allows for the condensate to be collected,  from which the rough rate of sparing can be calculated and also the rate of loss of peroxide. Because of the pressure drop through the dry and wet gas lines, the keg is under quite a bit of pressure and not realising this I unplugged the gas in line which caused peroxide to spray all over me. I was wearing full protective equipment it wasn't an issue but it does go to show how easily mistakes can happen. Because of that we decided to let the wet gas come out the large opening in the top of the keg so that it is not under any pressure. Bellow is a picture of me wearing my has mat suit with the setup inside. The keg and heater is inside the drum, which acts as a containment vessel in the event of a leak. Our safety protocol calls for wearing suits for bulk handling and pvc raincoats/waders with long gloves and face shields for collecting samples. We had one other incident when filling the contender where the cap wasn't porpelly on the peroxide container and some spilt on my suit. 

One thing I did not anticipate was how to deal with the vapour which comes off. I was under the impression that peroxide, like water didn't have any smell and that as long as we had good ventilation we would be fine. Well after about an hour of sparing in the workshop we all felt quite ill and had bad headaches so we had to move the setup to the doorway and have a fan to blow out the fumes. I think the strong smell comes mainly from the stabilisers