LEO Progress: RS-25 Adaptation

I love dictionaries (yes, I know, you’re shocked; shocked!).  I have several at home and at work including a two-volume abridged Oxford English Dictionary (OED) that was a wonderful gift from my mother several years ago. 


The definitions and word origin above comes from the OED on-line site.  Embedded within our language is so much condensed history and accumulated knowledge that it’s amazing.  While I have no doubt that this is true of every language, I only know my own to any significant degree.  Indeed, I’ve been babbling my own language for a long time now — forty-some years — but I can always pick up a dictionary and learn something new with just the flick of a page or two.  You really can’t say that about too many other things.

As the title for the article and the definitions above suggest, for this article we’re going to talk about “adaptation,” specifically about adaptation of the RS-25 engine.  As part of the Space Launch System Program, we are undertaking something a bit unusual for the world of rocket engines.  We are taking engines designed for one vehicle and finding a way to use them on another vehicle.  Now, this is not a completely unique circumstance.  The Soviets/Russians really were/are masters of this kind of thing.  But for us, it is not something that we do very often.  In terms of the big NASA program rocket engines, I can only think of the RL10 that was originally part of the Saturn I vehicle as an engine design that has had a long second and even third career on other vehicle systems.  The truth is that engines and vehicles are, for us, generally a matched set and the reason is that we just don’t frequently enough build that many of either.  Note that, technically speaking, we are also adapting the J-2X from a previous program, Constellation/Ares, but that’s obviously a bit different in scope and scale given where we are in the development cycle.

The name “RS-25” is, as I’ve mentioned in past articles, the generic name for the engine that everyone has known for years as the “SSME,” i.e., the Space Shuttle Main Engine.  There is much to talk about the RS-25.  Lots and lots of stuff.  More stuff that I could possibly fit into a single article.  Here are just some of the RS-25 topics that we’ll have to defer to future articles:

       •  History and evolution
       •  A tour of the schematic
       •  Engine control, performance, and capabilities

For this article, I want to just talk about the scope of work that will be necessary to adapt RS-25 to suit the Space Launch System Program. 



Clear Communication
The most significant thing that has to be done to the RS-25 to make suitable for the new program is that it has to be able to respond to and talk to vehicle.  Remember, the Space Shuttle was developed in the 1970s and first flown in the 1980s.  Yes, many things were updated over the years, but given the lightning-fast speed of computer evolution and development, it is not surprising that what RS-25 is carrying around a controller basically can’t communicate with the system being developed today for the SLS vehicle.  The SLS vehicle would say, “Commence purge sequence three,” and the engine would respond, “Like, hey dude, no duh, take a chill pill” and then do nothing (my lame imitation of 1980s slang as best I remember it).


But here are the neato things that we’ll be able to do: We can use almost all of the work now completed on the J-2X engine controller hardware to inform the new RS-25 controller and we can use the exact same basic software algorithms from the SSME.  Because the RS-25 has a different control scheme from the J-2X, we cannot use the exact controller unit design from J-2X, but we can use a lot of what we’ve learned over the past few years.  And, because we can directly port over the basic control algorithms, we don’t have to re-validate these vital pieces from the ground up.  We just have to validate their operation within the new controller.  That’s a huge savings.

This is work that is happening right now, as I’m typing.  Pratt & Whitney Rocketdyne, the RS-25 developer and manufacturer, is working together with Honeywell International, the electronic controller developer, on this activity.  Within the next couple of months, they will have progressed beyond the point of the critical design review for both the hardware and software. 

In the long term, it is our hope that we will evolve to the utopian plain of having one universal engine controller, a “common engine controller,” that can be easily fitted to any engine, past, present, or future.  Such a vision has in mind a standardization of methods and architecture such that we could largely minimize controller development efforts in the future to the accommodation of obsolescence issues.  The simple truth is that development work is always expensive.  It would be nice to avoid as much as that cost as possible.  With the new RS-25 controller, we’re getting pretty close to that kind of situation.

Under Pressure
Have you ever spent much time thinking about water towers?  As in, for instance, why are they towers in the first place?  As you might have guessed, this is something that I wondered about many years ago as a pre-engineering spud.  They seemed to be an awfully silly thing to build when you could just turn on the faucet and have water spurt out whenever you want.  Ah, the wonderful simplicity of childhood logic:  Things work just because they do and every day is Saturday.

The reason that you go to the trouble of sticking water way up in a tower is so that you have a reliable source of water pressure that can absorb the varying demands on the overall system.  In the background, you can have a little pump going “chug, chug, chug” twenty-four hours per day pushing water all of the way up there, but by having this large, reserve quantity always in the tower, the system can respond with sufficient pressure when, at six in the morning, everyone in town happens to turn on the shower at the same time to start their day.  The pressure comes from the elevation of tower, the height of the column of water from top to bottom.

On the Space Launch System vehicle we have something of a water tower situation except that, in this case, we’re dealing with liquid oxygen.  In the picture below, you can see roughly scaled images of the Space Shuttle and the Block 2 Space Launch System vehicle.  On both of these vehicles, the liquid oxygen tank is above the liquid hydrogen tank.  This means that for the very tall SLS vehicle, the top of oxygen tank is approximately fifty feet higher relative to the inlet to the engines (as illustrated).  This additional fifty feet of elevation translates to more pressure at the bottom just as if it were a taller water tower.  However, in this case liquid oxygen is heavier than water (meaning more pressure) and the SLS vehicle will be flying, sometimes, at accelerations much higher than a water tower sitting still in the Earth’s gravitational field.  Greater acceleration also amplifies the pressure seen by the engines at the bottom of the rocket.


“So what?” you’re saying to yourself as you read this.  After all, higher pressure is a good thing, right?  If you have more pressure at the inlet to the engine, then you don’t need as much pumping power.  So, life should be easier for the engines with this longer, taller configuration.  There are two reasons why this is not quite the case.

First, think about when the vehicle is sitting on the pad at the point of engine start.  The pumps aren’t spinning so all the pressure you’re dealing with is coming from the propellant feed system.  And now, simply, for the SLS vehicle it will be different than before.  One of the tricky things about a staged-combustion engine, in general, is that the start sequence (i.e., the sequence of opening valves, igniting combustion, getting turbopumps spun up) is touchy.  Given that the RS-25 has two separate pre-burners — and therefore three separate combustion zones — and four separate turbopumps, the RS-25 start sequence especially touchy.  You have to maintain a very careful balance of combustion mixture ratios that allow things to light robustly, but not too hot, and a careful balance of pressures throughout the system so as to keep the flow headed in the right direction and keep the slow build to full power level as smooth as possible.  We have an RS-25 start sequence that works for Space Shuttle.  Now, for the SLS vehicle, we will have to modify it to adapt to these new conditions.

The second issue to overcome with regards to the longer vehicle configuration and the liquid oxygen inlet conditions is due total range of pressures that the engines have to accommodate.  When sitting on the launch pad, you have the pressure generated by the acceleration of gravity.  During flight, as you’re burning up and expelling propellants, the vehicle is getting lighter and lighter and you’re accelerating faster and faster, you can reach the equivalent of three or four times the acceleration of gravity.  So that’s the top end pressure. 

On the low end, you have the effect of when the boosters burn out and are ejected at approximately two minutes into flight.  When this happens, the acceleration of the vehicle usually becomes less than the acceleration of gravity meaning that the propellant pressure at the bottom of the column of liquid oxygen can get pretty low.  Momentarily, the vehicle seems to hang, almost seemingly falling, despite the fact that the RS-25 engines continue to fire.  Pretty quickly, however, the process of picking up acceleration begins again.  (In the movie Apollo 13, they illustrated a similar effect with the separation of the first and second stages of Saturn V.  The astronauts are pushed back in their seats by the acceleration until, boom, first stage shutdown and separation happens and they’re effectively thrown forward.  With the Space Shuttle and the SLS vehicle, however, the return to acceleration is not as abrupt as it was on Saturn V where they show the astronauts slammed back into their seats with the lighting of the second stage.)  The point is that the RS-25 has to accommodate a very wide range of inlet pressures while maintaining a set thrust level and engine mixture ratio.  While this has always been the case for the SSME/RS-25, the longer SLS vehicle configuration simply exacerbates the situation.

You’ll note that I’ve not talked about the liquid hydrogen here.  That’s because, as I’ve mentioned in the past, liquid hydrogen is very, very light.  Think of fat-free, artificial whipped cream.  Yes, the top of the hydrogen tank is much higher, but due to the lightness of the liquid, it doesn’t make much difference at the engine inlet even when the vehicle is accelerating at several times the acceleration of gravity.

Some Like it…Insulated
Look again at the pictorial comparison of the Space Shuttle and the Space Launch System vehicles shown above.  Do you see where the SSME/RS-25 engines are relative to the big boosters on the sides of the vehicles?  On the Space Shuttle, the engines were on the Orbiter and they were forward of the booster nozzles.  On the SLS vehicle, there is no Orbiter so the engines are right on the bottom of the tanks and their exit planes line up with the booster nozzle exit planes.  In short, the engines are now closer to those great big, loud, powerful, and HOT boosters.  We are in the process now of determining whether this poses any thermal environments issues for the RS-25.  Thus far, based upon analyses to date, there do not appear to be any thermal issues that cannot be obviated through the judicious use of insulation.

Other environments also have to be checked such as the dynamic loads transmitted to the engine through the vehicle or the acoustic loads or whatever else is different for this vehicle.  The point is not that all of these environments are necessarily worse than what they were on the Space Shuttle.  It is only that they all need to be checked to make sure that our previous certification of the engine is still valid for all of these considerations.

The Tropic of Exploration
So those are the three most obvious and primary pieces of the RS-25 adaptation puzzle: a new engine controller, dealing with different propellant inlet conditions, and understanding the new vehicle and mission environments.  Each of these pieces carries with it analysis and testing and the appropriate documentation so there is plenty of work scope to accomplish.  We are extremely lucky to be starting with an engine of such extraordinary pedigree, performance, and flexibility. 

Henry Miller once said, “Whatever there be of progress in life comes not through adaptation but through daring.”  It is our intent to prove Mr. Miller wrong in this case.  We will make progress by using the adaptation of RS-25 to enable the daring of our exploration mission.

Liquid Engines Extra — Introducing LEO

The following picture is a test.  Find me amidst the mess…

That’s right.  I’m the handsome chap with the snazzy specs.  See, I’m just oozing with exactly the vitality that you’d expect from your typical government-trained civil servant!  Well, okay, so maybe I’m not exactly Milton Waddams.  All personal resemblance and charm aside, I don’t actually have quite that much paper clutter in my office.  No.  Instead, I just have electronic clutter.  Indeed, if someone spent the time to actually print out all the stuff on my computer here at work, then we wouldn’t need a rocket to get beyond the moon.  We could just build a staircase from the resulting gargantuan pile of paper. 

Why is this the case?  It could be that I’m just a data hoarder.  Some people hoard clothes (that’s not me).  Some people hoard automobiles (that’s not me).  Some people hoard books (okay, that is me).  And some people hoard data.  All data.  All of the way back to class notes from Aerodynamics I (AERSP 311, junior year, professor Bill Holl — great teacher, great guy).  Yes, okay, so maybe I’m a little guilty of all that.  In my defense, however, I will simply say that there’s a lot that goes into making rocket engines and much of it is quite removed from the exciting cutting-metal stuff or the making-smoke-and-fire stuff.  And I get to stick my nose into much of it.  For this article, I am going to reveal super-duper, deep-and-dark secrets that they won’t even teach you in college.  I am going to tell you a little about …

…wait for it…

…project management.  Yes, it’s true:  I am going to invite you into our little piece of office space and I will do so to tell you about how the office has recently evolved.  Also, towards the end as a reward, I’ll give you an update of J-2X testing progress to date.

So, let me introduce you to LEO —

 

No, none of those LEOs.  Instead, let me introduce you to the Space Launch System (SLS) Program Liquid Engines Office (LEO).  The LEO is responsible for development and delivery of the liquid rocket engines for the core stage and upper stage of the SLS vehicle.  For more information about the SLS Program in general, I highly recommend the following site: https://www.nasa.gov/exploration/systems/sls/

(Oh, and I want to say something briefly about the term “liquid engines.”  Probably every single person who would bother to read a blog about engines or rockets or space travel in general knows that this is just a shorthand term.  No, we are not talking about engines made of liquid — although that would be really cool.  “Liquid engines” is a quick and easy way to denote “liquid-propellant rocket engines.”  In case I’ve disappointed anyone, I’m sorry.  If ever we are able to make an engine out of liquid, I promise to be the first to report it.  Probably the most far-out thing that I once heard was the suggestion to make a hybrid rocket motor using solid hydrogen and liquid oxygen.  I cannot even imagine what the infrastructure would be to make the use of solid hydrogen plausible, but you never know…)

For the SLS vehicle, the upper-stage engine is the rocket engine so near and dear to our hearts after several years of design and development and fabrication and assembly and test:  J-2X.  The core-stage engine is the RS-25.  No, the RS-25 is not a brand new engine.  Rather, it is the generic name for that workhorse of the last thirty years, the Space Shuttle Main Engine (SSME). 

At the end of the Space Shuttle Program, there were fourteen SSMEs that had flown in space on the Shuttle and that still had usable life remaining.  I’m not sure that everyone knows this, but rocket engines have limited useful lives.  I guess that most things do, but with rocket engines it’s often pretty short.  Think of them like cherry blossoms (popular motif in Japanese tattoos): amazingly beautiful and quickly gone too soon.  The stresses within an operating rocket engine are tremendous.  For example, the J-2X has an official, useful life of only four starts and less than 2,000 seconds of operational run time after the engine has been delivered for use as part of the vehicle.  No, the engine doesn’t crumble into dust after that, but based upon our certification strategy and on our analysis of margins, that is the official life for our human-rated launch system.  After that point, depending on the proposed usage and risk considerations, and based on the likely reassessment of our margins with the proverbial “sharper pencil,” we can and do routinely talk ourselves into longer active lives for engine hardware.  On the test stand, we can test the J-2X upwards of 30 times and for lots of run time, but that is a lower risk situation.  Nobody is riding the test stand into space. 

Thus, when you come to the end of a program and you have fourteen engines with remaining, usable life, then you’ve got one heck of a residual resource.  In addition, there was one SSME assembled and ready to go, but it never made it to the test stand or the vehicle.  So it’s brand new.  And, on top of that, there were enough leftover pieces and parts lying around of flight-quality hardware to cobble together yet another engine.  And, there’s more! (Yes, I feel like the late-night infomercial guy, “and if you call in the next 10 minutes you will get this special gift!”)  There are also two development SSMEs.  These are not new enough to fly, but they are useful for ground testing and issues resolution.  That means that there are a total of sixteen RS-25 flight engines and two RS-25 development engines available to support the SLS Program. 

However, before your excitement bubbles over, you have to understand that when you see a sign for “free puppies,” you probably shouldn’t take that whole notion of “free” too literally.  As in, well, not at all.  Yes, we still have an extraordinary asset in the residual RS-25 engines.  No question.  But, we have work to do to integrate them into the SLS Program.  In a future article, I will discuss the multiple facets of this work.  By the way, I cannot claim to be immune from the “free puppies” thing myself.  Meet Ruugie –

The Liquid Engines Office (LEO) was formed to manage both the J-2X and the RS-25.  This office will also manage other liquid rocket engines used to support the SLS Program as it matures.  It was decided from a project management perspective that it would be best to have one office manage both engines.  In this way, we can be more efficient by leveraging the expertise across various disciplines and components.  For example, do we really need two turbomachinery subsystem managers?  No, Gary Genge is our turbomachinery subsystem manager and in that position he can understand and evaluate the relative programmatic and technical risk across all of the various turbomachinery pieces under his purview.  If in some utopian future our office responsibilities expands to three or four or eight different engine development or production efforts, we would, in theory, maintain the same structure but provide Gary with the support necessary to effectively manage turbomachinery across so many activities. 

So, for LEO we have subsystem managers for Engine Systems (effectively systems engineering and integration), Engine Assembly and Test (also includes asset management, logistics, and operations), Engine System Integration and Hardware, Valves and Actuators, Engine Control Avionics, Turbomachinery, and Combustion Devices.  LEO is supported by a Chief Engineer, a Chief Safety and Mission Assurance (S&MA) Officer, Program Planning and Control (i.e., the business office), and Procurement.  Plus, of course, we have support from the engineering and S&MA organizations across the many technical disciplines.  The structure is really quite similar to how we’ve been managing J-2X for these past several years.  We’ve just expanded our responsibilities.

So, that’s LEO and I’ll be talking more about RS-25 and SLS in the future.

Now, while I’ve been off doing my little part to get the foundation of LEO solid, including refreshing and getting into place our prime contracts for both J-2X and RS-25, how has J-2X been doing?  Well, in short, J-2X has been just cruising along.  E10002 has gone through six tests on NASA Stennis Space Center (SSC) test stand A-2.  Below are a series of images showing what an E10002 start looks like if you stood in view of the flame bucket (which I would very strongly advise against, by the way):

First, all you see is the facility water being pumped into the flame bucket.  Then you can see the ignition and everything glows orange.  Then the whole flame bucket is filled with exhaust.  And, finally, the exhaust coming barrelling down the spillway and eventually engulfs the camera.  The final step is not shown since there’s nothing to see but solid whitish grayness.

Here are the stats on the six tests:
     • Test:          A2J022          2/15/2013          35 seconds duration
     • Test:          A2J023          2/27/2013          550 seconds duration
     • Test:          A2J024          3/07/2013          560 seconds duration
     • Test:          A2J025          3/19/2013          425 seconds duration
     • Test:          A2J026          4/04/2013          570 seconds duration
     • Test:          A2J027          4/17/2013          16 seconds duration

So the total accumulated time is 2,156 seconds.  Tests #22, #25, and #27 all experienced early cuts, but all three were instigated by different flavors of instrumentation or monitoring system issues or oddities.  The engine is fine and running well.  Some of the key objectives included gathering additional data about the nozzle extension cooling characteristics, additional samples of the turbomachinery design, and main chamber combustion stability trials.  Something else that we did for this test series is that we tested a very special fuel turbopump port cover.  Here’s a picture of it:

Now, port covers are not something about which one usually says anything at all.  What makes this one special is that it was made by using a process known as Selective Laser Melting (SLM).  That is a fabrication method that is somewhat analogous to “3-D printing.”  A long time ago, I wrote a blog article about a gas generator discharge duct that we made for component-level testing using this technique.  This, however, is an engine test and this small, seemingly innocuous, piece of engine hardware may be the humble harbinger of a revolution in rocket engine fabrication.  The fact that we systematically stepped through the process of validating this port cover as a piece of hardware for an engine hot fire demonstration paves the way for pursuing other parts in the future, more complex parts, and, hopefully one day, regular production parts as part of a human-rated launch architecture. 

E10002 was removed from NASA SSC test stand A-2 on April 30th.  It is currently being retrofitted with instrumented inlet ducts and other hardware in preparation for the next phase of testing that will occur on NASA SSC test stand A-1.  As you’ll remember, in the past A-1 was used for the PowerPack Assembly testing.  Well, the talented and productive folks at NASA SSC remodeled the stand back to the configuration for engine testing.  The current plan is to install E10002 into A-1 by mid-May and to perform a series of five to seven tests through probably August.  The reason for using A-1 for the next series is because that stand does not have a diffuser.  That means that we can gimbal the engine, i.e., twist it around as if we were providing steering for a vehicle.  The thrust vector control (TVC) system composed of the hydraulic push-pull actuators that will be performing the gimballing is a component belonging to the stages element of the vehicle.  This testing will be providing those folks with data to inform their system design for the SLS Program.  See, it’s all win-win when we play nicely together.

And, finally, right on the heels of E10002, the assembly of E10003 will commence in June with scheduled installation into NASA SSC A-2 in September.  That’s my report for where things stand.  To finish up, I’ll leave you with a purely gratuitous glamor shot of the J-2X.  Isn’t she pretty?