After a long career in space and rocket program development, Salvatore T. “Tory” Bruno was named chief executive of United Launch Alliance in August 2014. In this new position, Bruno faces enormous challenges.
Over the past half-decade, SpaceX has emerged as a viable competitor, starting to fly its Falcon 9 rocket more often, and successfully competing for lucrative military launch contracts. Meanwhile, ULA was faced with a mandate from the US government to end its reliance on the Russian-made RD-180 engine for its workhorse rocket, the Atlas V booster.
In response to these challenges, Bruno has sought to reduce costs (through layoffs and other restructuring) and increase ULA’s commercial competitiveness, while developing the brand-new Vulcan missile with US-made components. Bruno also has to answer to two demanding parents, Boeing and Lockheed Martin, who each own 50 percent of his company and have competing aviation interests. Working for a little over four years, Bruno seems to be making progress. Most notably, the company recently won a $967 million contract from the U.S. Air Force to complete development of the Vulcan rocket, which ULA says will be ready to fly by 2021.
With this in mind, Ars Bruno spoke at the SpaceCom conference in Houston in late November. The resulting interview took so long that we split it into two parts. The first half, shown below, will focus on developing Vulcan and other parts of ULA’s business. The second half, to be published on Wednesday, looks at Bruno’s relationship with Blue Origin, which makes ULA both customer and competitor, and the company’s turbulent past and present with SpaceX.
arse technique: If you had to eat a hat, would you rather have it with mustard or mayonnaise?
Tony Bruno: I would have to say only with mustard as hats will probably be a bit dry and chewy.
But seriously, do you think you’ll be flying a national security payload onto Vulcan in 2022?
Yes I do. So our certification flights will be in 2021 for the LSA awards that have just been created. And we plan to fly a real customer on that flight, commercial customers, and we’re talking to them now. And then a few months later we are certified and ready to fly national security space missions.
What do you see as the long pole to get to Vulcan-Centaur’s first flight?
In a classic rocket development program, there are two things that are always the hardest things to get done, and they are always software and propulsion. And I expect this will be no different.
The propulsion seems to be making some progress with the BE-4 engine testing. But talk to me about the software. What’s the challenge there?
I think people don’t think of rockets that way, but they’re robotic vehicles. There is no pilot. They are fully automated. They have to be extremely fault tolerant because a rocket is a very complicated machine, and sometimes the different parts just work differently, or have problems, and the rest of the vehicle has to compensate. And because no human intervention is possible, a little different from even a spacecraft, because once it’s in orbit and spinning, usually has an A side and a B side, it can go into safe mode. We don’t have that luxury. There is little redundancy. And the whole mission is over in 15 minutes. The software has to be perfect.
And so people think, well, it’s software. If something comes too late, just change it. Thank goodness it’s just software, because changing the code only takes a few minutes. While physical hardware can take months to build.
It’s actually not like that at all. You have to have such a structured and disciplined process around your software development that it takes time to get to your final version, and if you go through a change you will go a long way upstream doing retrograde testing and qualification of all that . The software is incredibly complex. [With] every rocket I’ve ever developed – and it’s over three dozen systems now – software is always difficult to make because of the complexity of the process you have to use to ensure its high reliability.
You overtake a common avionics system from Atlas and Delta. Does that help you?
It definitely helps us because there’s a whole bunch of that software that will do the same thing. Depending on how you count it, anywhere from 40 to 60 percent of the software we use today will be usable as it is on Vulcan. So it’s a smaller amount of work to get done.
From a hardware standpoint, the biggest challenge is integrating new engines into a new core phase?
Yes, of course. They are two engines instead of one, complicating the fluid controls to bring the propellants from the tanks to the two engines. So it’s different from what we’re doing now, and everything has to be very weight efficient. We also have a brand new type of thruster construction, which I’ve shown pictures of, and some other parts of the rear end construction that I haven’t revealed yet that are super cool and use really new technology. So that’s all a very complicated systems engineering task to make that work on the gigantic scale that the rocket exists.
You also upgrade the Centaur’s top stage for Vulcan.
The very first Centaur we fly will be called Centaur 5. It will already have twice as much propellant as Centaur 3. Centaur 3 (which flies on the Atlas V rocket) has a diameter of 3.8 meters. The very first Centaur we fly on Vulcan goes straight to 5.4 meters in diameter. Then what we’ll do in the second upgrade to Centaur is upgrade the thrust in the RL10 engines and make it even longer, to stretch the propellant tanks to give it that much more energy. That’s what brings it all the way to the Vulcan Heavy configuration.
How is the RL10 upgrade paid?
There is currently a public-private partnership RPS contract developing the RL10CX, which also incorporates additive manufacturing into that engine. [Editor’s note: The RL10 engine is manufactured by Aerojet Rocketdyne.] The RL10 is exquisite. It is a Swiss watch. It is handmade. There are small tubes and passages. No excess beef or weight or anything on it. So to be able to put that into an additively manufactured engine is a huge benefit in time, cost and quality because once you put those automated processes in place and run them right, you’re not subject to the variability of people building them.
When we first fly the base Centaur 5, it has an existing RL10C on it. The other part of the Vulcan Heavy configuration, however, is that the tanks will not only be larger, but will also be powered by an additive-engineered, higher-thrust RL10CX engine. It involves a private investment by us and our partner Aerojet Rocketdyne, as well as a government co-investment. They now have an RPS contract to do that work. There will be work to finish it and integrate it into the vehicle that will be part of our LSA business.
You weren’t talking about ACES, the Advanced Cryogenic Evolved Stage. Is that dismissed as a top-level option as you try to get Vulcan out the door?
It’s just further in the pipeline. It has always been on our roadmap after flying Vulcan. It is still. And now that we’re in the hot and heavy development and manufacture and qualification of Vulcan, that’s what we’re talking about. But yeah, it’s still on the roadmap, further afield, where it’s always been.
Europe is developing a new launch vehicle. Japan is developing a new launch vehicle. SpaceX continues to grow. Blue Origin is working on a new rocket. In light of all this competition, how confident are you in ULA’s ability to increase its market share for commercial satellites?
I have a lot of confidence. I will be honest with you. We do very few commercial satellite launches these days. And what we’re looking for is modest growth and a relatively modest share of that market. Just as not all missiles are created equal, not all missions and not all customers are the same. What we see in the commercial market is a large segmentation. There is a large center of satellites and other vehicles that are fairly price sensitive, but they still balance price and risk in performance. There is another segment where it is all about the price. It is a relatively simple spacecraft; they’re able to insure its value, and their particular business model means they’re in no rush to get into orbit, so they’re willing to take planning risks to get into orbit — and even suffer a loss, so as long as the insurance covers it.
Then there are these guys here, who are very similar to the US government in terms of their risk tolerance. They have big, typically geocommunication satellites that are big, big capital investments. They don’t have many. Their business model and the economics of that market still support that. They can’t afford to lose one. They are usually in a hurry to hit the market. Since they are capital intensive, that’s another way of saying they spend the most money upfront, and it’s imperative that they get it into orbit in time, to start a revenue stream to close that .
For many of them, in that group, they think seriously about things like direct-injection spacecraft. That is an ideal marketplace for us. So we’re going to have good success there, but we’re not going to have success here (Bruno waves to the market segment of customers most interested in price, and chuckles), and we’ll see how we do in the middle.
I haven’t heard much about the XEUS lunar lander concept. Did you stop because Lockheed is chasing its own lunar lander?
It just fell off the priority list for us. So now we don’t do anything with it. We’re tired of getting Vulcan ready. And make Centaur 3 to 5, and then the bigger one. And then ACES. XEUS is so far away that it’s actually off the map.
But that was a very interesting piece of your Cislunar 1000 architecture. Presumably you want to come back to it someday?
I don’t know, it depends if other people do. So we’re looking at this whole transportation network, and where we’d fit in and add the most value. If other people take on that specific role, we don’t have to.
You’ve talked a lot about the commercialization of the moon. It seems that this administration has really embraced the idea of the moon as the next step and a commercial path.
Yes. I wonder how that happened.