November 8, 2003 notes

 

1000 lbf Mixed Monoprop

 

Safety note: we found a crack in one of our 5 gallon HDPE carboys that we use as a transfer container when loading peroxide from drums to our run tanks.  We always keep the carboys on a containment palette, so there was no spillage, but a few hundred ml of peroxide has probably leaked out into the containment sump when I noticed the crack.  We pumped the peroxide to a new container and tossed the cracked one out.  I checked my records, and we had been using that particular carboy for a bit over two years, which was probably a mistake.  The first year of that was with 90% peroxide, which is known to embrittle PE after extended exposure.  We had a bunch of new carboys, but we hadn’t tossed the old ones out.

 

Our main task this week was getting more thrust out of the large nozzle engines.  Because we were still getting the “cold hole in the center” of the chambers when the catalyst wasn’t packed so tight that it choked the flow down, we decided to try a pair of catalyst blocks in a straight-through configuration with a spreading plate at the top, instead of the three-pass plates on top and bottom.  We had tried this exact configuration in the past, but we had reason to believe that the addition of the perforated metal flame holder under the three-pass pack was the important point that made everything work.

 

We put this new top chamber on the same bottom chamber we built last week.  It seemed like it was more difficult to get preheated from a completely cold state than with the three-pass top pack, but after it was warmed up it made an very smooth run.  The chamber walls and nozzle heated up completely uniformly, and the thrust trace was very smooth, with less than +/-5% thrust variance.  After the run, the catalyst was completely uniformly hot across the entire surface, with no cold center.  Unfortunately, it didn’t make much more thrust than the previous runs.  We repeated the test with identical results

 

289 lbf average thrust during the run with 230 psi tank pressure, 342 lbf at 230 psi at line clear.

 

The thrust bump at the end of our test runs is due to having a 12’ –10 hose going from the propellant tank to the engine.  There is normally a fluid pressure drop when the line is full, but as the last of the propellant starts to go through the line, there is less fluid resistance, until the last of the propellant is just being pushed through the valve, with no effective hose length.  We often consider this “line clear” thrust rating as what we would get on a vehicle with very short, high flow plumbing from the tank to the engine.  All reported thrust numbers are also 10 – 15 lbf on the low side, because the load cell calibration has a negative zero bias at the moment.

 

Not having to use the three-pass plates is a big fabrication help.  I was having a miserable time milling down a ½” thick piece of 316 stainless plate for the big motor, and I wasn’t looking forward to making a ship-set of the small ones.

 

We cut open the hot chamber and removed the top section of catalyst bale, leaving two separate sections with 330 grams of catalyst bale pressed to 1.5” height each.  The results were another smooth run, but still hardly any increase in thrust after adjusting for a slight difference in tank pressure:

 

327 lbf average during the run with 248 psi tank pressure, 375 lbf at 250 psi at line clear.

 

We made another run at higher pressure, which was even smoother than the previous runs, with only +/-1.5% thrust variance.

 

507 lbf average with 393 psi, 555 lbf at 393 psi at line clear.

 

http://media.armadilloaerospace.com/2003_11_08/smoothRun.gif

 

This is actually good enough for us to fly the big vehicle for initial tests, but still lower than we expected from the engine.

 

We still don’t like working with the catalyst bale, because it is difficult to build engines consistently with it, and it obviously compresses down a whole lot under pressure, causing self-restriction.  We wanted to try building a hot chamber with only the foil monolith catalysts, which are consistent and have very low pressure loss.  We had tried that a lot before, but without the cold section / flame holder arrangement.  The 1” thick, 5.5” diameter 400 cell-per-square-inch foil monoliths weigh 275 grams each, so two of them will by 16% less catalyst than is in the current chamber, but will flow a lot more.

 

We were out of fresh monoliths, so we started scavenging from previous experiments.  We cut open the welded-together top chambers with the two previous three-pass catalysts, and we were pretty surprised to see that the outsides of the catalyst rolls were pretty chewed up.  The original three-pass pack had been propane preheated with initial start by torching the outside of the chamber, so that was understandable, but the second one had only had cold propellant through it, so we had expected them to come out almost perfect.  It looks like the widely varying temperatures in the different pass sections, combined with the plate bars restricting radial expansion, result in the foil pores at the intersections getting a lot more bent than unrestrained single pass catalysts.

 

We picked two monoliths from an old hot chamber that were the best of the lot, but they were still domed a bit, and had ragged edges that didn’t fit very tightly.  We separated the two monoliths with a perforated plate with 3/16” square stock on each side, in hopes that it would act as a flame holder section between the two blocks.  We left about a 2” section for the glow plug ignition chamber between the cold catalyst and the hot catalyst.

 

The assembly with used components was sort of ratty, but it did turn out to work, and made a good deal more thrust than the old chamber:

 

448 lbf at 253 psi tank pressure, 537 lbf at 252 psi line clear.

 

We made a longer run with similar results:

 

425 lbf at 248 psi tank pressure, 503 lbf at 237 psi line clear

 

This was a solid 30% more thrust than the old pack, but we were seeing something odd – the runs stayed perfectly clean, with no sign of quenching, but things just weren’t getting as hot.  The nozzle wasn’t glowing as much, and the inside of the engine after a run was only dull red instead of bright orange.  Calculated Isp on a later run was also down, working out to a measured 116, so even counting some lost to the warmup pulse (not much on the hot engines), we were down 15% or so from the good runs.

 

Increasing pressure yielded similar results:

 

616 lbf at 392 psi tank pressure, 755 lbf at 391 psi line clear

 

A longer repeat test gave the same results.

 

We increase the pressure again, and it continued to run clear with increasing thrust:

 

911 lbf at 618 psi tank pressure, but the nitrogen bottle and regulator couldn’t keep up with the flow, so the line clear had dropped to 927 lbf at 539 psi.

 

We increased again, and cleared 1000 lbf:

 

1010 lbf at 700 psi tank pressure, 1066 lbf at 672 psi line clear

 

http://media.armadilloaerospace.com/2003_11_08/highThrust.gif

 

The restrictions on this engine are probably in the plumbing and spreading plate now, not the catalyst.

 

This all looks pretty good.  If we fill the big vehicle tank to 400 psi, we can get 3000 lbf max from the four engines, and vehicle + 500 lb of propellant will be about 2000 lb.  Giving some margin for differential throttling, this will be a slow, dramatic liftoff, but it can go for 25 seconds.

 

We hope the Isp comes back when we build fresh engines with the new 600 cpsi catalysts that are arriving next week, and we are going to go ahead and add a third catalyst to the hot section to help out, but even if we don’t get any better than 120 Isp, we can still do a lot of flight testing.  We will do another set of engine development tests to optimize before building a ship-set of the 12” diameter engines for the X-Prize flights.  We will need our launch license before we can really fly the vehicle in that configuration.