At the heart of the WISP system is the ADDAPT SoC: Agile Distributed Data Acquisition, Processing & Training System-on-a-Chip. Analog Devices Inc (ADI) - as a global leader in the design and manufacturing of analog, mixed signal, and digital signal processing integrated circuits (IC) - has integrated all the wafer-level components necessary to facilitate an intelligent distributed diagnostic platform and packaged them in a 10 x 10 x 2 mm chip that weighs less than 0.5 g (~1g when fully encapsulated). Each ADDAPT SoC boasts 3 analog differential inputs, 6 Kelvin connection measurements, and 6 configurable digital communication inputs (SPI, I2C, UART or GPIO). An embedded microprocessor can be programmed to locally filter, fuse or otherwise process signals (including machine learning algorithms) to efficiently transmit only high-value data down the serial bus with both low-speed (1-wire) and high-speed (RS-422) options available to accommodate mass and power constraints. At least 64 devices can be daisy-chained over 30+ meters with a mix of sensors to support on- or off-line multi-system health management across an asset.
-
Witness Integrity Sensor Platform
The Witness Integrity Sensor Platform (WISP) is a novel sensor architecture developed in collaboration with Analog Devices Inc. (ADI) for facilitating condition-based maintenance (CBM). WISP was developed with the philosophy of minimizing complexity for the end-user. Novel sensors that are easy to install with simple diagnostic outputs that avoid complicated algorithms. State-of-the-art daisy-chainable hardware that is as small and lightweight as possible. Options for wireless power and data transfer to minimize cables and connectors. WISP is meant to be compatible with both legacy and emerging assets, with unlimited flexibility to accommodate both monitoring and integration challenges. The WISP system has been thoughtfully designed and rigorously tested through commercial and defense standards for mechanical loading and environmental exposure, and has been field tested on both ships and aircraft.
-
Fatigue Crack Gauge
Multiple advanced ADDAPT-compatible sensors have been jointly developed with MIT. The majority of these sensors incorporate polymer nanocomposites (PNC) within microfabricated assemblies to create passive sensing elements that permanently encode damage-related information physically, even when completely unpowered. The most mature of these PNC gauges is the WISP Fatigue Crack Gauge. The nano-scale elements behave as a network of resistors connected in series and parallel with each other, and surface-breaking cracks tear these bonds causing the network resistance to increase proportional to the crack length. A formal detection sensitivity study has determined that the a90|95 value (crack for which the sensor has a 90% probability of detection with 95% confidence) for the WISP Fatigue Crack Gauge is 0.32mm. While the current gauge is being produced with an active area of 12.5 x 12.5mm, nearly any size or shape could be produced-including ones with internal cutouts-at a mass of 10 mg/cm2.
-
Corrosivity/Erosivity Gauges
Leveraging the mature PNC crack gauge design, sensors for evaluating the corrosive and erosive potential of a hyper-local environment have been developed. Both gauges operate using the same principle. The conductive PNC is plated with an exposed witness layer—stainless steel for corrosivity or silver for erosivity—which electrically shorts a section of the PNC network. As the witness layer degrades in the presence of the offending environment, the balance of the parallel resistance between the PNC and metal plating shifts, thereby increasing the overall gauge resistance. The useful lifespan of the gauge can be adjusted by increasing the witness layer thickness. These gauges are calibrated using military standards for accelerated corrosion and erosion to obtain curves for damage metric versus exposure time. Subsequently, by following the same standards on customer materials to measure damage versus exposure time, one can obtain a mapping between damage metric and true damage. Such gauges can efficiently aid operators in predicting the potential for material degradation due to accumulated exposure to harsh environments without the need for power or continuous data recording.
-
MD7-Pro Digital SHM System
In-service monitoring techniques utilize a dense web of analog sensors connected by individual wires routed to centralized data acquisition and processing units. This traditional approach carries a significant weight penalty, can be complex to instrument and is susceptible to EMI. To address these issues, MDC has developed a fully digital SHM solution. The MD7-Pro system is composed of 3 core elements: an Accumulation Node for remote data concentration and diagnostic processing, an Acquisition Node for distributed signal digitization, and analog sensor bases that mate with both types of nodes. Each element of the MD7-Pro system is networked on a 10-wire serial bus that carries command, data download, node synchronization and power. Benefits of this distributed infrastructure approach include higher fidelity data through digitizing sensor signals at the point of measurement, reduced computational burden through local signal processing and feature reduction, and overall minimal mass through the consolidation of cables and elimination of bulky centralized hardware. The MD7-Pro System has been flight tested on fixed-wing aircraft, rotorcraft and spacecraft.
-
Accumulation Node (MD7P4-ACC)
The Accumulation Node is the first element placed at the front of any MD7-Pro bus. Measuring 60 x 40 x 5 mm with a mass of 25 g, the fundamental role of the Accumulation Node is to serve as an interface between the SHM network and the platform being monitored. It accepts 28VDC to distribute power for 24-254 daisy-chained Acquisition Nodes in a MD7-Pro network, along with relaying commands, facilitating synchronization, and storage of the resulting data. It can be programmed to run autonomously, communicate over Ethernet, or accommodate flexible provisions for other wired and wireless protocol. In addition, the Accumulation Node offers 16 digital input channels and provides 8-GB of static memory. A powerful FPGA with an ARM core processor can be programmed to execute embedded diagnostic algorithms or prognostic logic.
-
Acquisition Node (MD7P4-ACQ)
The MD7-Pro system can be efficiently expanded by daisy-chaining Acquisition Nodes. Measuring 50 x 40 x 5 mm with a mass of 20 g, the Acquisition Node is a direct replacement for traditional instrumentation such as rack-mounted oscilloscopes and function generators, enabling distributed data acquisition and signal processing. Each Acquisition Node provides a 20 Vpp 20 MSample/sec arbitrary function generator, 6 independent 12-bit channels of up to 50 MSamples/sec with programmable gain up to 500 or attenuation down to 1/500 in addition to 2 Gbit of DDR3 memory. The nodes are potted in urethane to provide resistance to moisture, chemicals, flame and shock loading, and have been designed to pass aerospace EMI standards. The Acquisition Nodes are capable of facilitating high sampling rate damage detection methods (guided wave, acoustic emission, and frequency response) or collecting multiple external pre-conditioned differential voltage sensor signals (strain, humidity, temperature, acceleration, etc).
-
Structural Sonar Array
The MD7-Pro Acquisition node can accept a large range of appropriately configured sensor bases, which not only provide a mounting interface to the structure, but provide a physical connection to the external analog and/or digital sensors being monitored. Traditional SHM methods require dense sensor meshes to precisely resolve the position of damage, however this drives up system weight and complexity. Thus, in addition to customer specified custom configurations, MDC has patented a standard PZT beamforming array package to facilitate both active and passive structural sonar scans. The Structural Sonar Array can indicate damage event coordinates using a novel sensor design along with an innovative algorithm with minimal information about the material or structure. From a single node position, a probability of damage map can be generated in response to stiffness changes detected by an active guided wave scan, or due to the passively captured acoustic response from an impact event. Results from multiple nodes can be combined synchronously and/or asynchronously to enhance sensitivity and resolution.