November 15, 2003 notes

 

Air Liquide Peroxide

 

We fired one of our engines with 50% unstabilized semiconductor grade peroxide from Air Liquide.  It worked perfectly.  This actually looks cleaner than the 50% propulsion grade we got from FMC – the Air Liquide reads 1 ppm on the TDS meter, while the FMC stock reads 4 ppm to 5 ppm. (stabilized Solvay food grade read 18 ppm, and stabilized FMC technical grade read 232 ppm)

 

It looks like our supply problems are finally over.  We have a solid engine combination, and a willing propellant supplier.  Hurray!  It sucked that the lack of 90% peroxide prevented us from flying any vehicles for the last eight months, but all the work that went into the mixed-monoprop propulsion system has probably been for the best anyway.  The propellant is several times cheaper, gives 20% better Isp, and is generally better to work with (less caustic, and less material sensitive).  The development work has also given us catalysts that we will be able to use with 98% peroxide when we are ready to work on high energy upper stage biprop engines.

 

Misc

 

The back-to-back tests we did with the FMC peroxide and the Air Liquide peroxide were with the older catalyst-bale hot section, which still heated up and glowed as hot as ever, so the high-flowing monolith hot sections are definitely not getting as efficient combustion.  We hope that the denser catalyst in the new batch fixes that.

http://media.armadilloaerospace.com/2003_11_15/2003_11_15_f.jpg

 

We tried a simple experiment this week to see if it was possible to dissolve some gasoline into our propellant by mixing 15% gasoline with 85% methanol (the M85 alternate fuel mixture), then mixing it with the peroxide.  There was some hope for it, because alcohol does allow some water to dissolve in gasoline, which is normally insoluble, but when mixed with peroxide, the gasoline separated right out.  We wanted to get some gasoline in for one of the same reasons it is put in M85 – to give a visible flame.  A visible rocket plume is useful to see some aspects of the engine behavior, and our perfectly clear runs don’t give us anything to look at.

 

We welded up a third large 4-way distribution manifold for the big vehicle.  We worked out the process to make leak-free aluminum welds, which involves using the pulse mode on the TIG welder to get better cleaning action, and really thoroughly cleaning the welding rods, which had more crud on them than we thought.  Unfortunately, we dinged one of the 37 degree AN sealing faces on the manifold, which is really easy to do after welding, because the aluminum has lost much of its temper.  We tried making a little gasket for the sealing cone, but that didn’t fix it.  Phil worked out a little lapping tool to clean up the face, and I think we have it sealed up now, but we may go ahead and make a stainless steel manifold.  I’m not going to bore a 4” deep blind hole in a block of stainless, so we are probably just going to weld stainless AN fittings onto a capped off 2” pipe section.

 

We tried to get our really expensive S-mass coriolis mass flow sensor plumbed up on our test stand so we could get accurate dynamic Isp measurements, but it seems to have developed a problem after sitting in a corner for the better part of a year.  Luckily it is still just barely under warranty.

 

We have been going through so much nitrogen with all the engine testing that we have moved to getting pre-manifolded bottles delivered.  We are going to get a liquid nitrogen dewar soon to start pressurizing the big vehicle tank, because it would take nearly a full six-pack just for a single test.

 

http://media.armadilloaerospace.com/2003_11_15/2003_11_11_a.jpg

 

We also replaced all of our carboys with brand new ones this week, throwing out the other ones that have been in use for the previous year or two.

 

Welded Engines

 

We needed a 5.5” ID engine tube to tightly fit the catalysts for the production vehicle engines.  We couldn’t find any 5.5” ID 316 SS tube on short notice, so we had some chambers rolled, welded, and bored true.  This was surprisingly expensive, the 10” long rolled tube cost more than the CNC machined-from-bar-stock nozzles when we had a decent size batch made.  We knew the wall thickness would vary a fair amount after machining, so we went with ¼” thick plate, even though our existing chamber extensions have been 1/8” thick tubing.  Square / cube scaling laws will make fabricated-from-plate options cheaper as we go to larger engines, but even at 12” diameter, CNC machining is competitive with metal forming options.

 

http://media.armadilloaerospace.com/2003_11_15/2003_11_15_b.jpg

 

These engines are going to be completely welded together, and will mount to the vehicle from a top flange, rather than a middle flange like our previous designs.  It will look a little odd having the bolt holes still visible on the welded-on-nozzles, so we may chop off the protruding part of the flange with the plasma cutter and smooth it back down with the welder.  The mill-finish on the chamber and top flange make it a rather redneck looking rocket engine compared to our all-CNC ones, but we are expecting it to work great.

 

http://media.armadilloaerospace.com/2003_11_15/2003_11_15_m.jpg

 

We used our plasma cutter and circle-cutting attachment to cut out the top flanges from ¼” thick 316 plate, and we cut a whole stack of 5.5” diameter perforated metal plates for the various flame holders and separators.  The plasma cutter is an incredibly useful tool, especially for working with stainless steel, which I loathe cutting on the milling machine.

 

http://media.armadilloaerospace.com/2003_11_15/2003_11_15_j.jpg

 

 

We have had enough problems with the laser cut spreading plates warping back up to the tops of the engines and disrupting flow that we are welding standoff blocks on each plate before welding into the engine.  The ones around the outside are just gauges to let us get the plate welded in the correct height below the top of the tube.

http://media.armadilloaerospace.com/2003_11_15/2003_11_15_d.jpg

 

The chamber, nozzle, and top flange together weigh 17 pounds, which could safely be cut in half if we cared to save the weight.  All the catalyst and internal plates will probably add another 5 pounds, so the engine T/W ratio is around 50, depending on how much pressure you put behind it.

 

Unfortunately, our big order of catalyst monoliths did not arrive this weekend, so we didn’t get to actually fire one of the new engines.  With any luck, we will be able to put together the full ship-set of four engines next week.  It actually looks like the big vehicle is going to lift off before we fly the small one again, unless we finally get some 90% peroxide from Don Stark in the next week or two.