As commercial aircraft generate huge caches of structural and system health-monitoring data, opportunities for predictive inspections and repairs are expanding. In tandem, sensor technology is further evolving as a key component in the process.
The short-to-medium-term goal is to be able to automatically identify all types of structural damage caused by fatigue, corrosion or accident during regularly scheduled maintenance intervals as defined in the aircraft’s MRBR (Maintenance Review Board Report)/MPD (Maintenance Planning Document), according to Johann Bordais, vice president of services and support for Embraer Commercial Aviation. “By replacing traditional inspection procedures such as general and detailed visual inspections—and non-destructive inspections such as X-ray, eddy current and dye penetrant—less time would be needed to perform time-consuming disassemblies to gain access to inspection areas,” he says. “The inspection times would come down to minutes instead of hours in some cases.”
Bordais cautions that most applications of structural health-monitoring (SHM) systems for commercial aviation to date are not yet being accepted as an alternative for traditional maintenance inspections. However, he notes, there are ongoing R&D initiatives.
“Embraer is already benefiting from the large use of SHM sensors in our full-scale fatigue tests, in terms of reduction of tests’ costs and times,” Bordais explains. “Also, the results of our SHM flight-test campaign, and the feedback from the airline participating in it, indicate that we are on the right track to develop a solution that will bring positive results in terms of structural inspection cost-reduction.”
Various entities involved with aircraft sensor technology are working at enhancing data management, rather than simply viewing the data, reports Amrita Kumar, executive vice president of business, at Sunnyvale, California-based Acellent Technologies.
“What you want to know is what is the data telling you, and how can you manage that information efficiently?” Kumar says. By moving toward on-condition—instead of scheduled— maintenance, “you do not have to disassemble the structure to do an inspection, due to the more targeted sensor technology available today.”
Kumar adds that while the presence and location of a problem is detectable, the next challenge is to characterize the damage. This, she says, will become increasingly important with the expanding use of composite materials.
“Is the damage in the form of a disbond or delamination? While sensors can detect that the damage is there, we’re still working on technology that will tell the operator what the exact nature of the damage is,” she says.
Acellent, reports Kumar, has a “draft version of the preliminary classification software” that would be applicable to composite structures. The application is in testing at the beta level, she says. Once tested, the software would be part of the company’s SHM Composite damage detection software, as an enhancement. “The SHM Composite software will detect, localize, size and classify the composite damage,” she says.
Ravi Rajamani, engineering director of Meggitt’s engineering and technology department, points out that delamination is considered “a big issue with composite structures,” since it can lead to structural weakening and failure. “Unlike metals, visual inspection cannot readily determine if a composite structure is compromised,” he says. “Embedded sensors, consisting of piezoelectric elements, are being proposed for failure detection of composites.”
In this regard, Rajamani reports that the biggest challenge presented is the extent of “the real estate” that needs to be covered. “Unless one can predict, a priori, where the damage is likely to occur, covering the entire superstructure is nearly impossible because of the weight and the cost of the sensors and electronics,” he says. “However, if we can identify vulnerable areas, then just those areas can be monitored, and this makes it more cost-effective.”
One possible approach to structural monitoring, he says, is distributed optical sensing. “This has a lot of potential, because fiber-optic sensing techniques do not need individual sensing elements. This makes such a sensor ideal for large structures, especially to measure temperature and strain.” On the downside, the associated electronics can be bulky and expensive, which is why, Rajamani explains, it will be a while before they transition from research to aerospace applications. “They are far more common in the oil and gas industry and in monitoring civil structures such as buildings and bridges, where there is not as much of a premium on low weight and compactness.”
While, as Rajamani states, there may be fewer sensors dedicated to prognostics and health management (PHM) on an airframe, compared to engines, “it is generally agreed” that all sensors serve a PHM function and “can be leveraged for implementing diagnostics and prognostics” capabilities. “The limitation often has been data bandwidth and computing power. That made it prohibitively expensive to actually get the data transferred to an appropriate location for analysis,” he reports. “But this is changing in a hurry, and we see increasing use of existing data to develop analytical solutions to help the operations and maintenance functions.”
Jim Berg, GE Aviation’s chief engineer, controls and accessories, reports that the industry, for engines, is moving toward more “continuous data flow” processes and away from “discrete data,” which provides a piece of information captured at a given time but nothing beyond that. “This is the result of a continuous process of engine health monitoring and sensor technology development that has taken place over the past 20 years,” Berg points out.
Darin DiTommaso, GE Aviation’s general manager of digital services and solutions engineering, adds that the OEM’s progression toward big-data analytics has enabled it to bring large data sets together, including information about environmental conditions, and configuration and operational usage. “That, coupled with data science and engine physics, is executed on a contemporary software platform called Predix, a cloud-based GE product that provides the capability to handle vast amounts of data for each engine in an efficient and scalable fashion,” he states.
Using the GEnx as an example, DiTommaso says that the widebody engine is fitted with 20 sensors, providing data for monitoring temperatures and pressures, fluid systems, fuel and oil, vibration, and valve settings and positions. The sensors are provided by GE as well as outside suppliers.
“The industry is in a better position today to leverage data from existing sensor suites, thanks to advances made in data processing, storage and data science combined with expertise to detect and isolate engine health issues and to leverage more of the data than we were [able to] in the past,” DiTommaso says. This means that the industry is moving toward more planned than unplanned engine maintenance, reducing inspections and the overall maintenance burden.
Sensors, of course, transmit data over an onboard cable network. Using engines as an example, the sensor collects raw data on the condition of the engine and sends it to the health-monitoring unit, which is integral to the full-authority digital engine control (FADEC). The FADEC does some internal processing and analysis of the data it receives from the sensor, which can result in real-time alerts. In addition, there is some unprocessed data that can be downloaded to a ground- based system via multiple methods—either during or after the flight. Because the wiring adds weight, research is ongoing into wireless sensors.
For example, GE Aviation’s DiTommaso says wireless data transmission for sensors is among the new technologies the OEM is researching for engine health monitoring. Among the benefits he cites are greater reliability and weight savings.
“Another advantage of wireless would be to reduce operator burden associated with a manual download of the data,” he notes. “The data becomes available much faster and in a more complete set than is typical with manual downloads. Also, wireless frees up the technicians for other activities.”
Berlin-based Beanair Wireless Systems’ co-founder Maneli Parsy says that wireless sensor systems may supplement, rather than replace, conventional data transmission by cable.
“The biggest problem is the increased volume of onboard transferred data within the aircraft or spacecraft, which requires a greater volume of cabling,” Parsy explains. “This is why, in most cases, wireless solutions will be looked into as complementary—not competing—with wired solutions. Other converging factors such as data compression and storage, miniaturization and the development of smart, distributive systems will also play an important factor in reducing the onboard data volume exchange.”
Asked about the potential for improved data accuracy using wireless sensors, Parsy cites the company’s SpaceWireless technology.
“Its main task is to challenge existing technologies by introducing a reliable, ultra-low-power and time-synchronized wireless network particularly adapted to dynamic measurement,” he says. “In testing, a significant improvement in accuracy was achieved through improved data transmission reliability.”
Parsy reports that SpaceWireless technology is undergoing tests with Airbus and Zodiac Data Systems. The main target of the testing is the deployment of a wireless sensor network—based on the technology—for flight-test measurement use. But its deployment on a production aircraft is not out of the question.
“SpaceWireless is already deployed on Airbus Flying Test Platforms, but as needs and systems evolve, we do not exclude design, test, certification and further embedding of wireless sensor networks within everyday operating aircraft,” he says.