Robot NDT said to scan structures in record time

 - January 11, 2008, 10:50 AM

A system that robotically inspects the skin and immediate interior structures of an aircraft–which has been likened to an MRI of the human body–is catching the attention of the aviation industry, including the Defense Department, at least one fractional-ownership provider and safety agencies.

The structural anomaly mapping (SAM) system will reveal corrosion and other subsurface aberrations such as debonds and composite delaminations without first having to do any preparation to the aircraft itself, according to the manufacturer. The aircraft remains in flight-ready status, with neither defueling nor contact with the aircraft coatings required.

Using a laser-guided vehicle, SAM can inspect the entire airframe of a Gulfstream IV in approximately eight to 10 hours, depending on the resolution of the scan. The results are stored in a database and can be used for comparisons in subsequent inspections.

“The second a measurement is taken it can be compared with every other measurement of the same spot on the same [airframe] or of any other [airframe] of the same type that has ever been measured,” explained Clark Freise, v-p of strategic development for Advanced Power Technologies (APT), which is working with Honeywell Engines, Systems and Services on bringing the SAM to market. “You can instantly start doing trending and fleet analysis.” That is something that particularly interests the military and operators of large fleet types, such as fractional providers, as well as safety agencies such as the NTSB.

In a demonstration shortly before Christmas, the SAM, which resembles a rolling oil derrick with a perpendicular boom, trundled around the hangar at Signature Flight Support at Ronald Reagan Washington National Airport, relying on pre-placed stanchions and a laser guide to position itself alongside an old stripped-down Hawker 600 airframe.

This sensor delivery system is controlled through wireless communications links with the SAM software, based on laser measurements and a digital 3-D model of the airframe. The articulated horizontal arm on the delivery device can move the sensor to examine complex surfaces, including downward- and upward-facing surfaces, from a distance no closer than approximately three feet.

Freise said that the prototype system was scaled to a GIV because BBA, the parent of Signature, was APT’s original partner. The gantry goes up to 30 feet and the arm extends to nine feet, “so there is no piece on a GIV we couldn’t image,” he added. The SAM was initially tested in Baltimore on Honeywell’s own business jets.

Automatic Fault Detection
While the delivery system has been built to accept a broad range of sensor systems, the prototype primary sensor is a laser velocimeter with adjacent acoustic source, which is able rapidly to scan large areas and automatically detect most faults on composite materials, subsurface frame damage and corrosion.

According to Honeywell, this performance has been demonstrated against both laboratory standards and portions of aging military aircraft. The algorithms used to detect faults are contained within the database coordination system, where structural anomaly measurements can be compared to library data for identification and classification. There is also “good potential” to detect skin cracks and widespread fatigue damage in aging aircraft with the current sensor, the company claimed.
Additional sensors (laser ultrasonic, thermographic, eddy-current or microwave) can be added or substituted in the future, since the entire SAM system and its software have been designed in a modular fashion.

3-D Digital Images
The heart of the SAM is a software system that coordinates the database, delivery system, sensors and processing software packages. It contains a 3-D digital model of the aircraft being mapped, which Freise called “one of the keys” of how this is a different methodology that brings information technology to these inspections.

The digital model is used to control the delivery device around the aircraft and to steer its sensor arm into proximity with the portion of the aircraft that is to be inspected. The database also holds information pertaining to previous similar inspections of the airframe and any other similar airframes in the same fleet. This allows for quick comparisons to previous measurements, so that detection of changes can be used to uncover anomalies in the airframe that may be caused by aging or unusual events such as bird or lightning strikes.

While most newer aircraft may have 3-D digital models that were used in their design, in the case of the demo Hawker, APT took the maintenance manuals and came up with a method to recreate them in 3-D. Freise said it takes about $10,000 per airplane type and a couple of weeks, which he termed relatively fast.

“That’s the worst case,” he said. “What we are finding more and more is that for most of these aircraft, somebody has already made a Catia model for maintenance or some other purpose.” The system is scheduled to be moved to Honeywell in Phoenix early this year, and Freise said that a Dassault Falcon 20 would be used to continue development there. In that instance, Dassault already developed a digital model for maintenance, even though the aircraft was not designed using Catia.

Customizable for Any Model

Thus, the mapping system can be customized for any aircraft, and it allows automated, complete inspections of airframes with quick comparison of results across a fleet of aircraft to detect early indications of fleet-wide problems. It not only reduces the manpower needed for inspections, its fully digital, repeatable methods remove human interaction, interpretation and variability from the equation.

That enables the system to make a precise measurement at the same point on the same aircraft every time. “It can do it a hundred times, a thousand times, it can do it over time, it can do it to different aircraft in the same fleet,” said Freise. “You get the same results every time. To do that we had to go into the fully automated system. Then we ran with the idea of removing the human elements, and going straight to digital measurements onto a digital model.”