December 2010 – Liquid Rocket Engines (J-2X, RS-25, general)

J-2X Progress: Turbomachinery: One More Pic

The previous article for this blog, the one down below regarding progress with the J-2X turbomachinery, was written about 2 weeks ago. There is an inherent lag in the process for big articles. That’s just the way things are. So, that being said, I am tossing a supplemental entry on here to provide an update: The turbopumps for Engine 10001 are assembled!

Below is a picture of the Pratt & Whitney Rocketdyne team responsible for assembling the Fuel Turbopump standing around the completed unit. Way to go guys!

Inside The J-2X Doghouse: The Gas-Generator Cycle Engine

Welcome back to the J-2X Doghouse. The last time that we met here, we discussed the fundamentals of what exactly makes something a rocket. As I explained, on the conceptual level, rockets aren’t really “rocket science.” You get the propellants together, light them on fire, and eject them out the back end of the vehicle. Simple enough.

Okay, but how do you move that much propellant and make that much smoke and fire, enough to propel something as big as, say, the Saturn V that was over 300 feet tall and weighed millions of pounds? That’s where things get interesting and technically difficult. As I said before it is all a matter of power. And to get power you use an engine.

What makes a rocket engine an engine is the fact that it contains more than just a combustion chamber where the propellants mix. It is an arrangement of machinery that, once started, feeds and powers itself. During operation, a rocket engine uses some cycle – some circuit of piping and thermodynamics and combustion and valves and control system and rotating machinery – to keep itself up and running and generating thrust.

Think about your car engine. You turn the key, the engine gets up and going, and then it can sit there for hours idling, running happily all by itself, converting gasoline and air into mechanical energy, with no additional input from you. You don’t have to manually pump the gas into the injectors (or the carburetor). You don’t have to plug it into an outlet to feed it more electrical energy. It’s self-sufficient until you turn it off or until you run out of gas. That’s what truly makes it an engine. It’s similar with a rocket engine except that the product is not mechanical energy; the product is very fast moving gases generating lots of thrust.

For rocket engine conceptual design, in terms of making it an engine, the goal is always, “How do you keep the pumps pumping?” These are extremely powerful pumps moving lots and lots of fluid, so you need some powerful energy source to drive them. The answer is to use what you’ve already got in the engine: the propellants. There are different ways to do this and thus you have different engine “cycles,” i.e., component arrangements. The most common rocket engine cycles are the gas-generator cycle (examples include J-2X, J-2, F-1, RS-68, and Vulcain 2 – see pictures above), the expander cycle (examples include RL10 and Vinci), and the staged-combustion cycle (examples include Space Shuttle Main Engine and RS-170/180). In addition to these, there are many other cycles and variations as well. Each different cycle has advantages and disadvantages and, usually, constraints linked to physics. Choosing the right cycle to fit the mission application is generally the first decision that an engine designer has to make. Because this is a blog dedicated to J-2X, I will focus on the gas-generator cycle engine.

Ideally, what you would want to do with a rocket engine is use all of your propellants in as efficient manner as possible meaning that you would want to use all them in the production of thrust. In a gas-generator engine, however, you concede right up front to a loss of some efficiency to achieve greater engine simplicity. You use a certain amount of the propellants brought into the engine almost entirely to keep the engine running rather than for generating thrust. In practice what this means is that you have a separate, small combustion chamber within the engine that does nothing but produce gases to drive the turbines connected to the propellant pumps. As compared to the large quantities of propellants being pumped through the whole engine, the amount going to the gas generator is small (less than 3% for J-2X), but once used to drive the turbomachinery, the exhaust is drained of much of its thrust-generating energy.

Below is a simplified schematic of a gas-generator cycle rocket engine like the J-2X. The propellants, liquid hydrogen (fuel) and liquid oxygen (oxidizer), enter the engine and go immediately into the pumps: the fuel turbopump (FTP) and the oxidizer turbopump (OTP). There, the mechanical energy of the spinning pumps is turned into high pressures in the liquid propellants.

After exiting the pumps, a small amount of each propellant is tapped off to supply the gas generator (GG). The GG is, in essence, a small rocket engine embedded within the larger rocket engine. It makes hot, high-pressure combustion products, steam and gaseous hydrogen, that are used to drive first the turbine connected to the fuel pump and then the turbine connected to the oxidizer pump. After driving the two turbines, this still-warm gas is used first to warm the helium flowing through the heat exchanger (HEX) that is used to pressurize the oxygen tank of the stage and is then dumped along the walls of the nozzle extension to keep that relatively cool. The video below is a component test of the J-2X GG performed at NASA MSFC. Even with the relatively small amount of propellant that the GG burns, an enormous amount of energy is released to drive the turbopumps.

The rest of the liquid oxygen coming out of the oxidizer pump, meaning that which is not going to the GG, is directed through the main injector and into the main combustion chamber (MCC). The main injector is analogous to a fuel injector in a car engine except that here it injects two propellants through hundreds of injector elements. The effectiveness of this injection and the mixing of the propellants are crucial for overall engine performance.

The hydrogen circuit after the fuel pump is more complicated. This is because the hydrogen is used to cool the nozzle and combustion chamber walls. The walls of these two components are essentially hollow. They contain hundreds of passages for the hydrogen to flow thereby keeping the walls from melting due to the extreme high temperatures of the contained combustion zone. After doing its job as coolant, the hydrogen is then directed through the main injector and into the MCC. Not shown on the diagram is the fact that a very small amount of the warm hydrogen gas is tapped off prior to entering the main injector and is routed back to the stage to pressurize the hydrogen tank (like the helium through the HEX on the oxygen side).

It is in the MCC where the mixed hydrogen and oxygen combust to make steam and residual hydrogen gas. The temperature of that combustion is approximately 6,000 degrees Fahrenheit and in the J-2X the pressure is approximately 1,300 pounds per square inch. These combustion products are then accelerated to sonic velocity at the converging throat of the MCC and then to supersonic velocities down the diverging nozzle and nozzle extension. As discussed previously, it is the high-velocity expulsion of these hot gases that produces thrust.

Note that the turbine exhaust gases dumped along the nozzle extension still generate some thrust, but not as effectively as the combustion products that are accelerated through the nozzle throat. This loss of effectiveness is the price that you pay for this relatively simple engine cycle. As a comparison to a more complex engine cycle, do a web search for the schematic for Space Shuttle Main Engine (SSME).

FYI, the other items denoted on that GG-cycle schematic above are the control valves: the main fuel valve (MFV), the main oxidizer valve (MOV), the gas-generator fuel valve (GGFV), and the gas-generator oxidizer valve (GGOV). These primary valves, along with several other minor ones, are used to control the engine during the start and shutdown of the engine.

So that’s how a gas-generator cycle engine like the J-2X works. As this blog continues and as we head towards testing next year, I will continue to report on the progress of the components that make up the engine.

J-2X Progress: Turbomachinery — The Rotating Components

It was once pointed out to me that most of a rocket engine really isn’t a whole lot more than a jumbled bunch of specialized plumbing. Notable exceptions to that general rule are the engine controller — the brain of the engine — and the rotating components, i.e., the turbomachinery. Of course, the person who was telling me this was a turbomachinery person, which means that I cannot entirely concede the point lest I yield my traditional posture of giving them a hard time. But there is no denying that rocket engine turbopumps are truly remarkable pieces of machinery.

What is a rocket engine turbopump? Typically, and this is true for J-2X, a turbopump consists of two parts: a turbine and a pump (hence the name). Pump portion is what draws in the propellants into the engine the pushes that fluid through all of the “plumbing” that leads, ultimately, to its fiery, thrust-generating expulsion. The turbine portion is what provides power to drive the pump. The turbine converts the power of hot gases into the power of rotational machinery. The pump converts the power of rotational machinery into fluid power otherwise known as pressure (thousands of pounds-force per square inch) within the propellant being pumped. For J-2X, there are two turbopumps: one for pumping liquid hydrogen (fuel) and one for pumping liquid oxygen (oxidizer, or often called “LOX”).

Soon, I will be writing an article for this blog that further explains the system-level workings of a gas-generator-cycle rocket engine like J-2X. So, stay tuned.

Recently, the Pratt & Whitney Rocketdyne (PWR) / NASA turbomachinery team has made significant progress toward completing the final assemblies of the hydrogen and oxygen turbopumps for the first J-2X development engine (E10001). The first two images show two major milestones for the liquid oxygen turbopump. In the first picture, the turbine-end manifold (top of the photo) is shown being mated to the pump-end volute that is secured in the build dolly.

J-2X Liquid Oxygen Turbopump after Successful Turbine Manifold Installation

The second picture shows that the oxygen turbopump has now been flipped over with the pump end now near the top of the image and the turbine manifold below. It is sitting in an oven where it underwent a drying operation after successful insertion of the first-stage turbine disk and the turbopump shaft.

J-2X Liquid Oxygen Turbopump Following First Stage Turbine Disk and Shaft Installation

The hydrogen turbopump has also made good progress by completing all pump-end assembly operations and the turbine manifold installation. The first picture of the fuel turbopump below was taken after the successful assembly of the impeller into the bearing support, and subsequently that bearing support assembly being installed into the pump end volute, which has been chilled in cryogenic liquid nitrogen. The nitrogen was used to create the proper fit for the volute and the bearing support to prevent hydrogen leakage under engine operating conditions.

J-2X Liquid Hydrogen Turbopump After Successful Mating of Volute and Turbine Bearing Support

The process of (1) chilling one metal piece so cold that it shrinks, (2) heating another metal piece so warm that it expands, and (3) then fitting the two pieces together in those states is a process used throughout engine assembly on many different components. It is a means for accomplishing an “interference fit” (also called a “compression fit” or a “press fit”), which means that the two parts, machined to their appropriate tight tolerances, would otherwise not quite fit together — almost but not quite. At room temperature, the pieces would interfere with each other if you tried to push them together. The chill/heat process during assembly allows them to fit together very, very tightly.

The second fuel turbopump picture below shows the successful installation of the turbine manifold onto the turbine bearing support representing a major milestone in the assembly process.

J-2X Liquid Hydrogen Turbopump Turbine Manifold Installed Onto Bearing Support

In the beginning of this article, I told you that rocket engine turbopumps are remarkable pieces of machinery. Yet, what I have shown you in the pictures are mostly images of shiny-metal external pieces, big hulking manifolds and volutes. For reasons largely having to do with export control considerations (Rule #1: blog author does not go to prison!), I cannot show you pictures or detailed schematics of the inner workings. I can describe them by saying that on the pump side you have an inducer, which looks like a fluid screw, and that feeds an impeller for a typical centrifugal pump. On the turbine side, I can tell you that there are two rotating disks of turbine blades and, effectively, two rows of stationary blades called stators or nozzles. And in between the pump ends and the turbine ends are a series of seals that separate the two ends. Ideally, the only contact between the pump and turbine ends would be the mechanical power of the rotating shaft.

To give you a better appreciation of the “remarkable” aspects of these units, let’s consider these machines in terms of their output. In terms of horsepower, the table below compares various machines with which you are likely familiar. At only 30 inches long and 20 inches in diameter, the J-2X hydrogen turbopump produces an incredible 16,000 horsepower. This power level is equivalent to more than 120 automobiles, or 90 light aircraft, or even 5 diesel-electric locomotives. In terms of energy generated in a small package, the J-2X fuel pump provides almost as much power as a large aircraft engine on the Boeing 747.

The two turbopumps for the first J-2X development engine are currently on track to complete assembly in December. These units will then be boxed up, shipped to NASA Stennis Space Center, and await engine assembly. So, the first development engine coming soon! And then, it’s on to testing!

Note that thanks are due to Jeff Thornburg, Upper Stage Engine Element Deputy Turbomachinery Subsystem Manager, for providing the largest portion of the technical updates and pictures that informed this article.

J-2X Extras: Rebuilding the Past

Several years ago, I was determined and ready to buy a new vehicle. I happened to be at my grandparents place at the time in upstate New York and my grandfather saw me perusing the local paper for dealerships making good deals. I told him that I was interested in getting a new pickup truck, something that I could use to go back and forth to grad school and carry all my stuff.

“Well, I’ll tell you what,” he said, “the best darn vehicle I ever had was a 1937 Ford Pickup. That thing just ran forever, it seems to me.” Then he winked, smiled, and added, “And, even better, I met your grandmother while I was driving that thing.”

So I went on down to the local Ford dealership and announced to the salesman wearing a plaid jacket and striped tie that I wanted to buy a pickup truck. My new best friend smiled a huge smile, shook my hand, and led me over to the part of the showroom dedicated to their latest line of beautiful F-150s.

“No, no,” I said, “I want to buy a 1937 pickup.”

“But we don’t sell used cars here, son, and certainly not classics like that.”

“I don’t want a used 1937 pickup,” I replied. “I want a new 1937 pickup.”

“There is no such thing,” he said, in an obvious state of confusion or maybe annoyance.

I scratched my head. “I don’t understand. I mean, you guys still have the drawings and such, right? And if you can build these big, shiny new things, then you can certainly go back and build something more simple, right? My granddad told me not to get suckered in. He said that I don’t need all these new-fangled bells and whistles.”

For the next hour, the salesman, named Pete by way (Pistol Pete, he chuckled to himself), tried to talk me into buying one of the current year models. He showed me everything, explaining with fast-talking expertise the dramatic advantages that his trucks had over the competition and even, he tossed in for me, far older models. But I was not totally convinced. Pistol Pete just shrugged and gave me his business card with a scribbled phone number for someone at their corporate headquarters who might be better able to help me. I left thinking that might have just lost the best friend I’d had for the last ninety minutes.

The next day, I called the corporate headquarters and tried to make clear what I wanted. I got bumped from department to department several times until I finally got someone named George seemingly willing to indulge me.

I told George what I wanted, but I also told him that I was impressed with what Pistol Pete the salesman had shown me. I said, “I’d really like to get that 1937 pickup with an automatic transmission, with overdrive, and cruise control. I would really like more speed and better handling. Better gas mileage too. Also, I’m thinking that I need more safety stuff, so I’d like that pickup to have air bags, modern crushable bumpers, and the latest auto glass. Plus, I’d like a bit more life and reliability, so building in that self-diagnostic system and computer would be good. And I read that some new body materials are less prone to corrosion, so build it out of new stuff.”

“So,” said George, who sounded perpetually half asleep when he spoke, “you want a 1937 pickup truck, with all modern features, built to all modern standards, with more performance, with better reliability, and with greater safety. Do I have that right?”

“Finally,” I said, “someone who understands! You’ve got it! That’s exactly what I want!”

“Right, then we will have to design you a new vehicle from scratch. That will take about two years of design effort, building a few prototypes, and then another couple of years of road testing and then certification from the federal highway authorities — which, by the way, will result in a bunch more changes unless you only want to drive it on Sundays and holidays like an antique car. Overall then I’m projecting that we’re talking, maybe, forty or fifty millions dollars as a starting point.”

“Huh? What are you talking about? That’s outrageous,” I yelled into the phone in dismay. “You built this thing 50 years ago, didn’t you? It didn’t cost that much then for goodness sake, even with inflation. Surely you’ve got the drawings just lying around somewhere, right?”

“No, actually we don’t,” said George in his tired monotone. “And even if we did and we had to build exactly a 1937 pickup, as it was built back then, it would be a project. To start, we would have to rebuild all of the tooling, recreate the materials we used back then, and reconstitute suppliers who have long since gone out of business. Add to that all of your new requirements and, well, you’ve got a whole new vehicle, right? So, basically, we’ll just have to start from scratch.”

I hung up the phone in an utter daze. Several weeks later, I bought a little Toyota pickup. I drove it for lots of years. I plan to someday tell my grandson that it was the greatest thing ever on four wheels and see where that leads…

One of the first questions that I got when I started this blog was why we didn’t just dust off the drawings of the old J-2 used for the Apollo Program and use that rather that launching into the J-2X development effort. Hopefully this little story provides a bit of insight by way of analogy. As we go along, I will tell you about the actual changes between the Apollo-era J-2 and the J-2X of today.