Thursday, October 30, 2008

Level 2 is hard

A lot of people have advised me to ignore the 180 until the 90 is in the bag. Its probably good advice, but I’d like to start doing some 180 second development tests when we go out to the site to test the 90 second vehicle. The trip to the test site is a significant portion of the time involved with a test and so I’d like to use that time for testing both if possible.  I did a very detailed model last night. I started with the following assumptions:

  • It needs to stay as an amateur vehicle IE total impulse < 200K lb/sec

  • I’ll use an off the shelf LR-101 as the motor

  • I’ll use the same AMS industries Spun 5086 aluminum hemispheres for tankage that we used for the 90 second vehicle.

  • I’ll assume that the LR-101 ISP starts at 200 at design Cp and degrades along the same slope as Cpropep down to the minimum Cp (Cpropep says LR-101 should give 230 Isp at design pressure) This discrepancy is probably because the LR101 is significantly over-expanded at sea level.

I included in the model:

  • Pressure Drops in the LR-101 Jackets, and Injectors

  • Isentropic Gas expansion in blow down mode.

Given these assumptions I could not build a 180 second vehicle. so I modified the assumptions:

  • Add one Carbon fiber SCI 602 presurization bottle and Regulator to use on the LOX side. Allowing us to  fill the Lox sphere  past the point where blow down mode would run out of pressure.

Given these assumptions the model says we can hover for 200 seconds.

This probably won’t work for the following reasons:

  • The motor is WAAAY over-expanded for the entire flight. We could use all the LR101 dimensions and injector and build a motor that has the same internal dimensions and lower expansion ratio. this would likely give us back some ISP, but we are no longer off the shelf.

  • The motor needs to throttle 4:1 and as the pressure drop in the injectors goes down that low then we probably don’t get good mixing.

  • The outcome is really sensitive to the initial loading conditions over/under filling the Lox tank by 5% causes a 10 second change in the hover period.

  • We need to get the same hydro test performance out the tanks as Armadillo is getting and our first tank was 25% low, we have not yet been brave enough to test the 2nd tank to the level we need.

  • It would be really hard to keep the mixture ratio matched exactly over a 4:1 throttling ratio without really good closed loop controls.

  • The LR 101 is hard to get lit.See the SDSU rocket at 1:42 into 

Some additional thoughts on the 180 second problem:

If I use different sphere sizes and thicknesses and put on lots of pressurant bottles I can get my detailed model to say we hover for 230 seconds, but the same issues identified above apply.

The 180 second level 2 is all about getting good mass ratio and ISP. For non-pumped systems the density of the propellants really matters. I’m using 1.1 as my Lox density 0.8 as my RP-1 density. I really think Peroxide would be better, 90% peroxide has a higher density than Lox 1.36, and it also makes up a higher percentage of the total propellant load so the portion of your propellant at 0.8 is lower. I’ve seen Density * Isp ^2 as a figure of merit (FOM) in SSTO studies.

Examples :

  • Peroxide RP1 at 200 PSI running at best ISP has ISP 215.6 and Density 1.246 FOM: 58K

  • Lox RP1 running at LR-101 mixture ratios and 200 PSI has ISP 219 and Density 0.993 FOM: 48K

  • Lox RP1 running at best ISP and 200 psi  has ISP 230 and density 1 FOM: 53K

You are much more likely to get the peroxide motor running at peak ISP to cool in regen mode as there is much more cooling fluid available with a much higher heat capacity. Its clear the LR-101 designers with an infinite budget did not choose to give up 10 points of ISP with out trying.

One LLC competitor that had a static display at the 2006 xprize cup, had a system with electrically driven positive displacement pumps using peroxide, liquid catalyst and a fuel. this insures constant mixture ratios across the entire throttling range and makes your FAA safety system really easy as you just have a power relay that drops power to the pump and all propellant flow stops. Your tank thicknesses  are set by minimum gage issues not pressure requirements.  If one developed this it would also be easily transfered to vehicles with more aerodynamic tankage than the big spheres. (Space here we come….)


john hare said...

I have thought at times about using nitrogen expander to drive a vane motor that drives a vane pump. Vane motor being an impact wrench and throttleable. I haven't tried those numbers in a long time though.

Paul Breed said...

If you have hydrogen peroxide around why bother with a nitrogen expander....

If you split the vehicle into 4ths it is almost exaclty the right size to use a single hydrogen peroxide jug as the tank and 600 sized RC helicopter motors and batteries as the pump drive....

A Rocket with no tanks and no valves....

The motors would be tiny....
0.4 inch diameter throat, 1 inch diameter exit 4 inches long....

Chris Phoenix said...

OK, this is a very stu... I mean uninformed question, but:

Referring to your comment a post or two ago, that the first thing to do is always to break down the peroxide: Why does this have to be true in a biprop?

What I'm thinking is that sufficient heat will surely break down peroxide. If you add fuel, then you certainly get heat.

In a rocket engine, the propellants mix and react as they flow toward the nozzle, so the burning zone would likely be flushed out, as the heat wouldn't travel upstream fast enough to decompose the peroxide in order to generate the heat...

But surely, with modern materials, you could put in a flameholder? If you held a flame near the top of the engine, couldn't that provide enough heat to decompose the rest of the peroxide?

Damien said...

The problem is to get the combustion started, without causing a hard start. If you decompose the peroxide first, you get "automatic" ignition thanks to the heat from the peroxide decomposition.

Chris Phoenix said...

Well, OK, so the tradeoff then seems to be between a catalyst and an igniter. I guess I can see how a solid catalyst would be more reliable - until it gets stripped or melted. A liquid catalyst would seem to require more mass than the igniter, especially since the igniter could (in theory) use the same peroxide and fuel feeds, with a tiny bit of durable catalyst.

Paul Breed said...

Google General Kinetics.
They have a number of peroxide papers there to read. Several of which are papers on thermal peroxide decomposition. One shows that if you add 15% un-decomposed peroxide to a stream of decomposed peroxide the stream travels 300+ inches before 99% of the un-decomposed peroxide decomposes. This is only adding 15% one would prefer the ratio to be the other way around. Same site has a paper on a 250Lbf H2O2 RP-1 thrustor with an after catalyst L* of 7" and 95%+ C* I think we want to decompose first.

One of the future experiments I will run is to support a bundle of small copper tubes inside insulating ceramic supports, Maybe even vacuum jacket around the tubes so they don't loose heat to the environment. Then heat them up to 700 or 800C and see it they thermally decompose peroxide and retain enough heat to do this steady state.

Chris Phoenix said...

There seems to be a big difference in reaction rates between monoprop and biprop post-catalyst injection.

According to p. 11 of the Decomposition paper, the decomposition time is exponentially dependent on temperature. Even going from 90% to 95% peroxide (monoprop) increases the decomposition temperature enough to decrease the decomposition time by an order of magnitude. At biprop temperatures, decomposition takes microseconds.

So I would expect that the combustion / fluid dynamics would be similar to any other case of liquid (non-cryogenic) fuel and oxidizer: once you vaporize the droplets (which should be typically fast at combustion temperature), you can assume that the oxidizer is chemically available.


Lili Wang said...

michael kors handbags
michael kors outlet
oakley sunglasses
coach outlet
hollister clothing store
abercrombie store
coach outlet
louis vuitton handbags
timberland boots
soccer jerseys
coach factory outlet
cheap ray bans
louis vuitton outlet
cheap oakley sunglasses
kate spade handbags
burberry outlet
louis vuitton
oakley sunglasses
michael kors outlet
retro jordans
coach factory outlet
christian louboutin
caoch outlet
cheap jerseys
gucci outlet online
gucci outlet
mulberry uk