ATK designs key avionics and control systems for NASA’s Space Launch System booster.
On March 28, 2012, at a test facility in Promontory, UT, NASA and ATK successfully completed the first test for NASA’s Space Launch System (SLS) booster program. ATK, an aerospace and defense products company, hosted the demonstration [designated Flight Control Test 1 (FCT-1)] to test key avionics and controls designed for booster ignition, nozzle steering, and booster separation. The test specifically focused on the subsystem’s ability to start-up, monitor, steer, and shut down an SLS booster-nozzle, thrust vector control (TVC) system.
The space shuttle's original avionics and TVC were designed in the mid-1970s, and both flew on the space shuttle’s solid rocket boosters from 1981 through retirement in 2011. When Paul Karner, ATK’s program manager for avionics and controls, began work on ATK's next-generation booster he was looking at 35-year-old avionics. Given that new electronics are typically rendered obsolete by the time they enter production, Karner stood before a tall order.
"The avionics system that flew on the old booster had parts that were well beyond their life from a standpoint of being able to find replacement parts," says Karner. "When we had an opportunity on Aries 1 to redesign the avionics system, we went to a more modular approach, similar to what you would see on an aircraft."
Many of the space shuttle’s avionics resided in a single, common controller box that weighed more than 200 pounds. After a redesign by Karner’s team, the boxes now weigh 45 pounds (at most), so if an immediate remove-and-replace issue arose, the part could be swiftly swapped on the launch pad in a turnaround timeframe that wouldn't hinder the launch.
"We upgraded all of the electronics," Karner recalls. "We went from an analog type of electronic system to a digital one, and took more of an aircraft approach from the ability to support remove-and-replace, and get the vehicle ready for flight, sooner."
Karner set out to design a common chassis for the six different types of boxes, each serving a unique functionality. After his staff had a common chassis, they designed 14 circuit cards that could be mixed and matched to produce the various functionalities needed from each box. For the circuit cards, Karner kept the design as common as possible in order to benefit from significantly higher quantity buys on such things as field-programmable gate arrays (FPGAs).
From the analog standpoint, each of the old circuit cards had a very specific function that was extremely limited in its abilities. The old design featured a lot of discrete wires coming out of the circuit cards to perform singular commands; for example, On/Off commands. By going to a digital system, ATK was not only able to shrink down the circuit cards, but also increase the capacity on each card.
"By using FPGAs, and being able to talk over a Mil-Standard 1553 data bus, we were able to communicate quite a bit of data," adds Karner. "With an analog system, if you’re over a discrete bind, you'd only be able to get a small set of commands."
ATK was able to decrease the count of circuit cards, overall weight, and board size, all while increasing the functionality and amount of data that could transfer over the digital data bus.
The increased data transfer has allowed Karner to power up boxes and perform a built-in test to check the health of each box before it talks to the rest of the vehicle. It also allows his team to continuously monitor the health and status of each box and subsystem.
"It gives you the confidence for flight and if you do have an issue, it allows you to be more rigorous and specific on your troubleshooting," Karner adds.
To narrow the gap between test and time-to-market, Karner and his staff utilized an approach that called for requirements in early phases of the avionics' lifecycle. They took the requirements and decompose them into early prototype circuit boards dubbed "brass boards".
"We call them brass boards, because it’s literally a piece of brass that you layout an initial circuit design on to prove out its functionality," says Karner. "As the requirements matured, we matured the circuit design. But as we were modeling off of an aircraft approach, we had functioning power use, line replaceable units (LRU) within a two-year time frame from when we first started our requirements definition."
It was an aggressive timeframe from a launch vehicle aspect, but Karner was benefitting from lessons both he and ATK’s entire design community had learned from experience in the aircraft industry.
The team also made sure to have system safety, which is composed of a reliability group involved upfront in the design efforts. When the avionics team was developing a circuit, the reliability group would take an independent look at it to verify that the design wasn’t mixing critical and non-critical circuitry together.
"If there is any one message that I have for [design engineers], it’s that involving safety and reliability early and often is a big benefit." According to Karner, the avionics and control staff saved significant cost by having safety and reliability involved from the beginning, because issues were caught while they were still on paper, long before they even started to lay them out onto hardware.
"Today, the name of the game in this business is affordability," says Fred Brasfield, ATK’s vice president whose role was to oversee the overall development of the next-generation booster. "During the space shuttle era, it was always important to be affordable, but it was not the driving factor that it is becoming today. Given the realities of budget challenges at the federal level, it’s really become a big deal. We’ve signed on to deliver the same sort of reliable and safe performance that our shuttles did for years at a much more challenging price, and we’ve met that challenge."
Safety and reliability wasn’t the only outside group involved in early design iterations. From the beginning, ATK wanted its supplier to be a partner in this new government contract. When the company went into the proposal phase of bidding for the avionics, it asked the potential suppliers to offer options that would help increase liability and decrease cost.
"We didn’t just downselect a supplier, but a partner," says Karner. "We provided the requirements, but we wanted to make sure that we had somebody buying in with us. Part of that was giving the supplier design authority, but also the responsibility and accountability on how they wanted to implement it based on their design suggestions." As the list of potential suitors was weaned to one, ATK partnered with L-3 Cincinnati Electronics, an established pioneer in space exploration and military/defense technology.
Matching Design to Maturity
The challenges posed during the design and development of the avionics and controls represented nothing unique to the application. As with any program in aerospace, automotive, or construction, Karner’s biggest hurdle was matching the design to the maturity level and requirements.
"The biggest challenge was not letting the design get too far ahead of the requirements as both the stage and vehicle continued to mature," says Karner. "It was making sure that we kept pace with the requirements as they matured."
The basic requirements in Karner’s world included igniting the motor, staging the motor, and controlling the motor. He was able to capture high-enough level requirements to protect his part of the project against scope creep, because his team designed the basic functionality right out of the shoot. Even if NASA had tweaks for future vehicles, the avionics are common enough to handle most changes in either firmware or software.
Inside of the LRUs, the equipment doesn’t run any software. The avionics and control system all run all firmware in a language called VHDL. "We stayed away from airborne software," says Karner. "We wanted to benefit from the firmware. We didn’t want to worry about the certain problems that go with software as our goal was to avoid overcomplicating these avionics."
For the FCT-1 test, no booster was in the stand — it is currently being fabricated and will be tested in 2013. Karner’s team went up to Promontory and integrated their new avionics side-by-side with the digital test system that has been at the site for several decades.
What was important to this test was the steering system (the TVC). Karner and his staff went through a full simulated countdown and ran the system. The avionics successfully sent the signal to arm the motor, which was then verified in the closed loop system, and then sent the signal to fire the motor. The avionics was programmed to do a typical duty cycle, which was an over test of its ability to move and steer. It then performed the shutdown of the motor and TVC system.
"The system didn't know that there wasn't a motor for this particular test," adds Brasfield. "The boxes and associated software were actually able to perform a test firing. It isn’t as spectacular to watch — there is no fire and smoke — but these huge actuators cycled in and out at the exact length, at the exact right time. And for those of us who like to watch actuators, it was a big deal because it worked the first time."
The success was an impressive feat given the issues that ATK has had in the past when testing with the dated system in Utah. While Brasfield contends that ATK has never had an issue on flight, he notes that some of the systems, which are used to control the ground test, consist of "old flight metal" — previously designed flight hardware, and ground support electronics that have been around for a long time.
"In our very first demonstration motor for the Ares program, we had a test delayed because that system didn’t talk to the computer properly," recalls Brasfield. " The computer did the right thing and shut it down. Having been involved in this project from the very start, watching it move exactly the way it was supposed to move was very exciting."