Official Armadillo Q&A thread

I'm not going to talk money, but if it had been me I would have taken the Armadillo route (even given hindsight). The result of the route (so far), is that Armadillo now have a load of experience in peroxide motors, quite a bit in LOX motors, but most importantly they own all their intellectual property, there are no other people they have to pay. Now, they could have come to an agreement with whomsoever to take over the IP of any manufactured engine - but I would expect that would cost more than the $1.5m quoted.

They also have the chance to make breakthroughs - they are working at the problem from (almost) scratch, without the 'shackles' of prior experience. Most good ideas come out of left field, not with experience. Remember, if a third party is contracted out to do engine work, then they are going to play it safe - they have a contract to meet after all, so why do anything risky (unless they are happy with broken contracts and bad publicity, although it never did EDS any harm...)

Just a 2p's worth.

James

There is an even simpler view at the things.
Having two groups with recent development experience in similar engines is way better for the greater good of the rest of the industry. Having twenty would be even better.
besides, Armadillo has done a great contribution towards getting to twenty. I know i have learned alot from just watching them do this.

And another view : outsourcing any development has its benefits and drawbacks in practically every situation. as a leader of the project, you have to weigh both pros and cons. I think Armadillo did the right thing for their project.

I am excited to see the discussion regarding the "what-ifs" of what we could have done different to win.

Unfortunately our project is easy picking for anyone to second guess design parameters, mode of operation, or outsourcing. I expect it, and do not resent it at all! Please keep up the critique, it is a healthy way for us to be able to answer these questions to ourselves as much as to the world.

After the fact, as obvious as some other solution appears on the surface, there is always some underlying problem that prevents that from being the correct course of action.

As an example, lets consider the XCor engine route.

I have all the respect in the world for XCor and their ability to design an engine that would do what we want. Lets just take a moment and see if that really would have guaranteed us a win this year...

Armadillo would have had to provide them with a specification right after the 2006 X-Prize cup to give them time to develop this engine. They do thousands of engine tests to make sure it does what it is supposed to do before releasing the design.

The price would have been in the neighborhood of $1,000,000.00 or more because it would have occupied that much of their time, and would have been somewhat beyond their current experience. By that I mean that it is a very deep throttling engine. X-Cor currently does not throttle their engines.

The pressure drop they have through their jacket and injector are likely so much higher than what we use that we would have to double the thickness of our tanks to support that blow down pressure. Of course we could implement a pistonless pump... But now we are talking about a completely different vehicle that we have to design. That's not bad, it's just not direction we wanted to go for the sake of simplicity.

Now we have to scrap our modular vehicle design, and the quad...

X-Cor finally finishes the design on the engine probably somewhere in August 2007. What has Armadillo been doing with all that time? flight testing without an engine?

Next, Armadillo does not own the intellectual property of the engine design and "if" it does not do exactly what we need it to do, we can't change it, XCor has to. Invariably, Armadillo would try to make the engine do something they didn't expect and it would have some form of failure. Now we need to figure out why it failed and how to fix it, but the Cup is only 2 months away by that time.

As you can see, it is not as simple as it seems.

Armadillo MUST own the engine design and develop that expertise in order to be competitive with the time we have allotted.

Say what you will about what our engine did at the cup, but outside of the hard start issues, that engine has the capability of throttling nearly 10:1, and has a very small pressure drop allowing us to run a blow down system. This equals a light weight system capable of significant flight performance. (Not bad for a bunch of amateurs working part time in a garage...)

We are looking at some other possible changes to make to the engines that also include sacrificing some of the isp to get better thermal performance. This opens doors to other materials possibly. We will see.

The comments I find most interesting and helpful from everyone that posts here are the ones that are solution oriented. The breath of fresh air is that nearly all of the posts have provided that to some extent, even if we have internally discussed those options and decided we did not want to go that way, it is still good input.

The bottom line is, Armadillo is seeking balance between simplicity and performance, ease of manufacture and repeatability. Functional, consistent, and reliable will happen for Armadillo. It is only a matter of time.

If we were to eliminate one parameter, our job gets easier. Deep Thottling. That provides many significant issues for the design because you are not only changing the engine thrust, but you are reducing the cooling capability at the same time to some extent.

Obviously we cannot do away with Deep Throttling. That is what a VTVL does. We can reduce it, and have redundant engines. (look for that in the future.) That does double the plumbing though.

Meanwhile, keep up the comments and help us look forward.

Thanks!

Great post Phil, it's really nice to see that there is appreciation on the forums!

I think it is an ideal place for people to throw their ideas across and have constructive comments back. It makes a difference to know that you guys are taking the comments on board! Good luck going into 2008! It looks like it's shaping up to be an exciting year for Armadillo!

Rob

I am curious to hear what you guys think the structural integrity effects of heating fuel soaked graphite are. Would placing and impervious coating on the outside of the graphite improve consistency from start to start or do you count on the minor cooling you get from fuel seepage? It seems that eliminating the seepage and moving fuel flow control to post-jacket would allow a robust and repeatable start-up sequence

Thanks for the post Phil

[edit] Oh and have you considered dropping a couple grand to get a few XCOR guys out to the shop to talk about the specific challenges of Methane engines. A consulting call. Knowing a bit more before you start can't hurt, and I highly doubt that Xcor would not profit from the discussion as well.

I'll say!
I heard Armadillo has an agreement to work on Methane engines for NASA. Going pro on us already? Can you give us more details on that?

This is a very interesting question because we have had the same concerns without definitive data to suggest anything one way or the other.

Really though, I don't think the structure of the graphite is affected significantly. The start sequence could be affected, but if it is a restart situation, the graphite is hot enough that the embedded liquid slowly boils out of the graphite as the heat soaks back. If it is a first time start, the graphite is already dry. We also do a spark check to burn off any residual in the chamber before ignition.

We have seen water seep through the graphite very slowly under pressure, and surmised that the alcohol would have some possibility of doing this to have some degree of transpiration cooling. When we measured the pressure difference from the jacket to the chamber, the pressure drop was so small that we didn't think that would happen to any measurable degree.

The biggest problem we saw in our testing with the graphite was oxidation that would put a groove in the throat. We don't think it was so much burning the graphite away as it was oxidizing it away.

This being the case, we are going to chrome plate one of the next chambers inside and out to see if that protects the graphite from that type of problem.

In the past we used a ceramic coating with some success to prevent the erosion but went away from it because we didn't see te problem very often, and usually only when we leaned out the mixture ratio, or had an irregular burn in the injector that we had to fix.

If the plating does what we expect, then that will be a fairly cheap addition to the process, and will also remove any potential issues with absorption of the alcohol into the graphite. The plating may actually have more benefit with Methane than Alcohol.

On the LOX Methane front, We are excited about the work ahead. We have spoken with some of the Xcor guys already, and they have been very helpful.

We have also received feedback from the guys at NASA regarding some of the problems they faced with tests that did not go as expected. In fact they shared some of their past information on hard starts they had seen with LOX Methane specifically. The first thing I thought when I saw that was, "What do you know, we aren't the only ones!"

Some of the theories for the hard start problems as it applies to us or NASA, or anyone for that matter, include:

1) Igniter torch running too cool in general so that the ignition is not immediate.
2) A large Igniter torch running too rich causing greater than anticipated chamber pressure during the LOX lead, forcing hot oxygen back into the fuel injector and into the jacket before the fuel hits the injector.
3) An igniter flame front disassociating from the face of the injector and causing ignition further down the chamber rather than right at the face of the injector. Excess propellant behind this front ignites with a boom.
4) Igniter flame running too hot, causing the ports of the igniter to melt away and lose igniter chamber pressure. This usually lights the first run fine, but has trouble lighting the second run because of the decreased igniter pressure.
5) Insufficient purge in the Fuel injector just prior to ignition. (See #2)
6) Residual fuel wicking into the LOX manifold and freezing on the surface until the igniter engages, causing a small mixture of Fuel and LOX inside the LOX manifold. This could result in anything from a small pop to a sizeable boom.

Some of these things we theorized before the event and have already implemented procedures to alleviate. We are going to adjust the procedure and the design more to alleviate some of the others based on new information from the X Prize cup.

You can anticipate quite a bit, but you can't anticipate everything. There always seems to be a surprise waiting somewhere. We just have to figure out ways to reduce the odds on the surprises.

That's my quota of posts for the next week...

You might want to take a quick look at some of the published research on ceramic thrust chambers that are transpiration cooled. They use SiC/C (silicon carbide on carbon fiber) instead of C/C due to the lower oxidation. SiC/C might be an option for your nozzles, though it might require finding someone new to manufacture them. Plating the graphite nozzle as you mentioned might work too and be easier to manufacture, though the different expansion rates of the metal and graphite might cause the metal to flake off under heat load. A thin coating of SiC is another possibility since it might bind better with the underlying graphite and have a similar expansion rate.

In general, transpiration cooling seems to result in increased thrust due to less weight at the expense of some isp loss due to the reduction of fuel pressure. DLRF has an excellent simulation study on this. While I guess Armadillo isn't in a position to make a radical change in its engine technology at present, transpiration cooling might be worthwhile considering down the road at some point if you ever do. The downside is you'd have to learn manufacturing of ceramic components if you wanted to do the chamber yourself, much more difficult than machining it out of metal. There are some suppliers that provide nano-based compounds that can be formed, but this is definitely an art in and of itself.

One of the things I find curious about a lot of the alt.space development is how few technical risks people are taking in terms of propulsion. Even simple, proven concepts like altitude compensation nozzles aren't being investigated. I guess maybe the technical risks are being taken in other directions, like throttability, which it seems from this thread as if nobody has really investigated in the past. Or maybe people don't see the need for trying new tech because the immediate target is suborbital and the performance isn't needed yet. Perhaps that's OK, it sure was a thrill watching the X Prize Cup video where the Module moved like a real vehicle, just like a hovering car or bus, instead of an artillery shell. I wish Armadillo could scale it up quickly, so that I could fly from SFO to Narita in 3 hours instead of 10, at a reasonable cost. But I guess I'll have to be patient and wait a few years yet for that.

Thank-you for your detailed response. I am sorry to see that you have reached your weekly quota as answers usually just lead to more questions. I can happily wait till next week.

I am wondering why the graphite at all now that you have the fuel jacket. As you mentioned the pressure drop from the jacket to the chamber is not significant. The strength of the graphite is not needed. (speaking as someone who has never built a rocket engine) A thinner metal walled chamber (chrome plated aluminum?) would have better heat transfer characteristics and could possibly be lighter and cheaper. Including an overall reduction of the engine diameter.

I am sure that you have your reasons but it seems that the graphite is a throwback from the carbon wound design.

Do you try to keep flow in the jacket laminar? trade off between cooling performance and chamber pressure.

I was also wondering about the graphite, it would seem that the cooling jacket would make it unnecessary as DanielW mentioned. I was also wondering if it was what was causing your engines to be temperamental from time to time.

Have you guys ever talked about going to an all metal design anytime soon?

Graphite severely reduces the cooling requirements of the motor, you can let the graphite glow red hot and it will be relatively ok, while you'd have to keep an aluminium motor around 100-200 degrees C for structural reasons. This means a much smaller cooling jacket which as you said makes the motor smaller and could help with the hard start issues armadillo is having but you need to greatly increase coolant flow velocity and that means higher pressure drop across the cooling jacket. As Phil has said this would be a problem with their current blow-down design.

I'm doing my thesis next year on reciprocating rocket propellant pumps so if I were to design a rocket similar to what armadillo are doing I'd probably go that route, allow lighter propellant tanks and gives you more propellant pressure to play with however I can certainly see why armadillo went the way they did.

Okay, one more quick one... I came in a bit early today.

We have toyed with altitude compensation nozzles, but that was with peroxide monoprop. It looked cool and different, but it was a little harder to make and it didn't buy us anything because we weren't going very high. At that point, it didn't make sense to continue with that design.

In the not to distant future, we will be putting things to a much higher altitude, and we are again exploring ways to compensate for reduced pressure at altitude. The bigger factor is probably how to handle the expansion of multiple engines off of the base of a very wide vehicle rather than a single engine.

You are mostly correct when you say throttlability is the technical risk we are dealing with. that is precisely why we use graphite, because it is so forgiving. It is much easier to commit to pulling out most of the heat than it is all of the heat. It also allows us to keep a smaller pressure drop in the cooling jacket and conserve the available pressure to convert to chamber pressure and thus thrust.

We have considered transpiration cooling, but it appears that will be better suited for an optimized engine later if at all. If we can get the graphite to work, or if we can move to an all metal design with film cooling it may not make sense to pursue the transpiration cooled ceramic.

Graphite is very forgiving thermally, and allows us a wide range of throttlability without any significant thermal issues. This makes for a nice development platform that is relatively low cost. We do not have to be too concerned about distribution in the jacket because of this.

We are considering reducing the volume of the jacket by half however, and we dont think the pressure loss will be much worse as it still represents a significant amount more area than the injector itself. Flow losses at the doubled fuel velocity through the jacket however could be an issue that increases the pressure more than we might expect.

Last night we discussed an all stainless steel engine with just film cooling, and what isp losses we could expect to see if we could make that work. We will probably give it a try, and that will take us back to an engine design very similar to the carbon fiber wrapped graphite we used at X-Prize Cup 06. That engine started very smooth and nice every time. Any guesses as to why? Yep, no cooling jacket fill delay. The fuel valve was as close to to the injector as the LOX valve was.

In the musings last night, John said, "It would sure be nice if we could weld the injector onto the chamber and not worry about o-rings or cooling." It would be a bit heavy, but in comparison to the graphite, the jacket and the fuel in the jacket, it might not be as bad as one thinks. Worst case, perhaps we can put a graphite nozzle in a stainless chamber and run with that. We will no doubt try that very soon.

Whe we did aluminum regen engines in the past, we were very close to making them work. They would work steady state for a while, but as soon as we started to throttle them, or if we ran more than 30 seconds with any variance in the run, they would have burn throughs. This indicated to us that it took much longer to get to a steady state thermal condition than most people think.

Some of the things we did to try and alleviate the burn throughs was to change the cooling jacket... and that worked, but when we really had that down, the restriction was such that we lost too much flow for a blow down system. In addition, we added a very small percentage of ethyl silicate, which worked rather well.

We also hard anodized some of the engines. That seemed to work, but on another "naked" engine we polished the inside of the chamber to mirror smooth without anodization, and it appeared to work even better than the anodized engine. That said, perhaps a chrome plated aluminum would be a direction to experiment with, but it will be later than sooner I think.

Much of that regen work was done on a throatless engine design that ended up serving us very well. The side injection we had was not very efficient though as we discovered later, and if we had it to do over, we would use a top injector and add a significant amount of film cooling. that combination could very well work, and be extremely cheap and easy to fabricate.

We did end up using a phenolic chamber eventually for XP Cup 05, and for development purposes, that was extremely forgiving and very easy to work with. The disadvantade was having to replace the chamber every flight, but still that process was reduced to less than 5 minutes as part of the vehicle turn around.

If anyone else wanted to experiment with a LOX/anything biprop, a throatless phenolic chamber is a very nice way to go. If you want a little more life out of the chamber, go with a carbon fiber overwrap graphite chamber. Just make sure to film cool it a bit. We have a video of one of those letting go on a high performance run that didn't have enough cooling...

Okay, that really is my quota!

Why not tungsten or tungsten carbide? Is it too hard? Is oxidation a problem at high temperatures?

Was reading along and started pondering about the cooling jacket and associated problems... What follows are some musings about ways to control that and whether it would be a help or a hinderance:

It has been mentioned that the throttling results in different pressure/flow in your cooling jacket - and this has been a technical challenge.

I know it would add some weight and another part to fail (both the natural enemies of a good engine, I know) - but what do you think of the idea of some kind of pressure regulator or "blow-by" valve just up-stream of the jacket?

My thinking goes like this: Allow "X" pressure through to the jacket (either a static value or some adjustable value, preferably via a simple and reliable method - not an active control), and any excess is divertted back to the tank or straight to the engine (injector/nozzles/insert-desired-input-path-here). It might also aid the responsiveness of throttling at higher values (since the liquid takes a shorter path to the engine if the cooling jacket is already "full" and the bypass takes a shortcut).

Would this help with your issues at all, or is this just a solution looking for a problem? I can see that it might add failure modes; and I don't know what would happen when the colder "bypass" liquid re-mixes with the expanding warmer liquid (gas?) that has been run through the jacket - could mess with your mixture ratios and result in other funkiness, I guess. I also have no idea how your heat loads change with throttling and how much extra flow is necessary to keep the engine from overheating...

Anyone with experience have any thoughts on this? I'm just an amateur aerodynamic engineer (and mostly with gliders - no engine required!), trying to come up with a way to avoid fluctuating so many engine variables at the same time while you throttle.

--Noel

If I understand you correctly, such a design would be ok for a pump based system where the cooling jacket is at a much higher pressure than the tanks, but not in a pressure fed system. The problem is that you want flow through the cooling jacket and then back into the tanks, however in a pressure fed system this is just like having a U-shaped tube coming out of the tanks and then going back in again, with no pressure differential between the cooling jacket and the tanks there would be no flow.

Looking at some of their pics I assume the fuel from the cooling jacket passes directly into the injector body, a design that would probably help is to have an outlet from the top of the cooling jacket connected to a valve before re-entering the injector. That way you could totally fill the jacket before starting the engine and both propellant valves would be the same volume away from the injector however this design has its own problems and I'm sure it's been considered.