A_T_I_2015_04_

AEROSPACETESTINGINTERNATIONAL.COMAPRIL 2015 |^55

Automating NDT z

of multiple NDT techniques, along with
the use of advanced signal processing, is
expected to increase the defect detection
rate and size precision by more than
20% while reducing the inspection
time from minutes or hours – or days
in the case of maintenance, which
would otherwise involve dismantling
the engine to gain access – to seconds.

INTACOM PROJECT
Research is not confined to Clean Sky.
TWI in the UK is a participant in the
WELDMINDT project, and the TWI
Technology Centre in Wales recently
completed the first phase of its
IntACom NDT automation project
with the production of a prototype
robotic NDT system.
Backed by Rolls-Royce, GKN,
Bombardier and the Welsh government,
the three-year project aimed to achieve
a fourfold increase in the speed of
inspection and analysis of geometrically
complex composite components. And
it has succeeded to the extent that the
industrial partners have opted to back
further phases of the project without
waiting for a further round of Welsh
government funding.
High raw material costs and labor-
intensive processes make composite
parts expensive to produce, and the

requirement to inspect every part

means that NDT can be a bottleneck.

Exploiting the properties of composites

to produce more complex parts

exacerbates the problem and demands

both faster and more advanced

inspection techniques. The alternative

is manual inspections that can add

variability and cost.

According to project manager Ian

Cooper, the research addressed three

areas. As well as automating the

inspection itself and applying

advanced PAUT, it uses advanced

software techniques such as assisted

defect recognition and scan display

management. Automation means areas

previously scanned by hand, with

results transcribed on to the part and

report drawings, are now scanned

automatically using scanning paths

generated from imported CAD data.

The heart of the automated

inspection system is an inspection cell

comprising two six-axis robot arms,

capable of working independently and

cooperatively. The robotic arms deploy

end-effectors carrying ultrasonic array

transducers that can scan wide areas

of the part in a single pass. CAD data

imported into the system is used to

generate scan paths using commercial

or in-house-developed off-line

programming software. Mousing over

the displayed CAD image enables the

user to select surfaces to be inspected

and assign tools to each surface, while

live imaging allows programming of

focal laws and other ultrasonic

parameters. The tools and components

are protected by detachable magnetic

holders. Once the part has been

defined, scanning is achieved by

simply selecting the part from a menu

and pressing the start button.

ULTRASONIC ARRAYS

The replacement of single element

transducers with ultrasonic arrays

enables curved surfaces, radii and

other difficult areas to be scanned at

much higher rates than was previously

possible. Where possible, large PAUT

probes were used to provide maximum

area coverage in a single pass.

Ultrasonic modeling was used to

determine the optimum inspection

techniques for difficult areas of

geometry. For example, curved arrays

were used to inspect internal and

external radii in a single pass.

Complex curvatures are addressed

by detecting the shape of the surface

interface and adapting the focal laws to

suit. This can be pre-programmed or

done ‘on the fly’. Other challenges such

“THE HEART OF THE AUTOMATED INSPECTION

SYSTEM IS AN INSPECTION CELL COMPRISING

TWO SIX-AXIS ROBOT ARMS”

LEFT: The Morfi

robot uses active

thermography to

examine the

chemically milled

pockets in 737

Classic fuselage

skin plates

ABOVE: Lufthansa

Technik has

developed a

prototype robot

to inspect Boeing

737 Classic

fuselage panels