35
A third generation robotic Eddy current
system for the inspection of gas turbine
engine components
aB s t r act
Implementation of damage tolerance requires the
characterization of nondestructive evaluation (NDE)
capabilities in terms of the probability of detection
as a function of aw size. Quantication of NDE
procedures is essential to provide both a condence level
in detection of required aw sizes and in establishing
periodic inspection/maintenance events as a function
of part usage. Gas turbine engine components (aircraft
engines) are highly loaded and critical the safe-life/
operation of an engine. In general, detection of small
aws is required to provide the maximum mean time
between inspection cycles and thus contribute to most
economical engine operation. Reliable detection of small
aws in engine components has been demonstrated
by precision eddy current inspection procedures.
e procedures were implemented using a precision
robotics scanner and various eddy current probes and
scanning sequences. e eddy current system elements
were integrated into scanning sequences to provide
a quantitative, fully automated inspection of critical
areas of engine components. e integrated system is
identied as the RFC (Retirement for Cause) system
and has been extensively applied to the inspection of
rotating gas turbine engine components throughout
the world. Since the introduction of the system in
1979, additional engine components and inspection
features have been added to provide support of
various engines types and to extend the life of aging
components. e most recent system improvements
have been an improved eddy current instrument,
improved robotics controller and added computing
and electronic communications capabilities. is paper
described the retirement for cause system management
principles; the RFC Eddy current system development
and application; and the recent improvements to
increase capabilities and reduce both system cost and
system operating costs.
Key words probability of detection, nondestructive
evaluation, aw, inspection, maintenance, gas turbine
engine, eddy current, robotics controller
C P
S K
C B
1
Carlos Pairazaman
carlos.pairazaman@uniwest.com
2 Sara Keller
sara.keller@us.af.mil
3
Charles Buynak
charles.buynak@us.af.mil
Tercera generación robótica por corrientes Eddy para la inspección de
los componentes de turbinas de motores de gas
Recibido: marzo 11 de 2017 | Revisado: abril 10 de 2017 | Aceptado: mayo 15 de 2017
| C | L,  | V. XX II | N. 23 | PP. - | - |  |  -
https://doi.org/10.24265/ campus.2017.v22n23.03
36
re s u m e n
La implementación de la tolerancia al daño requiere
la utilización de las capacidades de evaluación no
destructiva (END) en términos de la probabilidad
de detección como una función del tamaño del
defecto. La cuanticación de los procedimientos
de END es esencial para proporcionar tanto un
nivel de conanza requerida en la detección de los
tamaños de defecto como en el establecimiento de
eventos periódicos de inspección/mantenimiento
en función del uso de las piezas. Los componentes
de los motores de turbina de gas (motores de
aeronaves) son altamente complicados y son críticos
la seguridad y en la de vida de funcionamiento
de un motor. En general, se requiere la detección
de pequeñas fallas para proporcionar el tiempo
medio máximo entre los ciclos de inspección y así
contribuir al funcionamiento más económico del
motor. La detección able de pequeñas fallas en los
componentes del motor ha sido demostrada mediante
procedimientos de inspección de corrientes parásitas
de precisión. Los procedimientos se implementaron
utilizando un escáner de robótica de precisión y
varias sondas de corriente de Foucault y secuencias de
escaneado. Los elementos del sistema de corriente de
Foucault se integraron en secuencias de exploración
para proporcionar una inspección cuantitativa y
totalmente automatizada de áreas críticas de los
componentes del motor. El sistema integrado se
identica como el sistema RPC (Retiro por Causa)
y se ha aplicado extensivamente a la inspección de
componentes de motores de turbina de gas rotativos
en todo el mundo. Desde la introducción del sistema
en 1979, se han añadido componentes adicionales
del motor y características de inspección para
proporcionar soporte a varios tipos de motores y para
prolongar la vida útil de componentes viejos. Las
mejoras más recientes del sistema han sido un mejor
instrumento de corrientes parásitas, un controlador
robótico mejorado y capacidades agregadas de
computación y comunicaciones electrónicas. En
este trabajo, se describieron los principios de gestión
del sistema de RPC; el sistema desarrollado y la
aplicación de corrientes de Eddy para el RPC; y las
recientes mejoras para aumentar las capacidades y
reducir tanto el costo del sistema como los costos
operativos del sistema.
Palabras clave: probabilidad de detección, evaluación
no destructiva, falla, inspección, mantenimiento,
motor de turbina de gas, corriente de Foucault,
controlador de robótica
| C | V. XXII | N. 23 | - | 2017 |
C P - S K - C B
37
Introduction
Early gas turbine engines have a very
short operating life between overhaul/
maintenance cycles. Both economic
and operational readiness requirements
have demanded and every increasing
operating life (mean time between
overhaul). Structural deciencies were a
primary cause of failure. Improvements
in design, materials and maintenance
practices greatly increased the operational
life, but were limited by knowledge and
methods for anticipating service induced
initiation and growth of aws. e advent
and introduction of fatigue and fracture
mechanics as design and life- cycle
management tools oered substantial
improvements in life predictions and
substantial improvements in both safety
and operational reliability. A key to
implementation of fatigue and fracture
mechanics was the development of
quantitative inspection capabilities
(Berens, 1992).
In the United States, the USAF
Material Lab (Man Tech Division)
assembled a team of experts and funded
a program to develop a state of the art
inspection system in October of 1981.
is system is known as the Retirement
for Cause system and was awarded to
System Research Laboratories Inc. (now
VEDA Corporation). e initial program
challenge was to reliably detect small
(0.005 by 0.010 inch) cracks in parts
that were being returned for overhaul
or replacement. Without the inspection
system, parts that had exceeded their
1,000-hour design life were retired to
minimize the possibility of a catastrophic
failure in ight. When condence was
gained with the new RFC inspection
system, parts could be returned to
service for an additional 1,000 hours and
beyond with substantial savings in parts
replacement.
e RFC Inspection System
e RFC system was conceived as a
fully automated inspection system that
could be operated by mechanics in the
engine overhaul facilities. Extensive
subcomponent development and testing
were completed to assure that the basic
inspection capabilities could be met. e
subcomponents were then integrated
into the automated system and further
system level testing and validation
was completed before each inspection
sequence (scan plan) was accepted
for production application. ese
practices and disciplines have extended
to all subsequent system versions and
improvement packages.
e resulting RFC Eddy Current
Inspection System (ECIS) was a
computer -controlled eddy current
inspection station with standard
communication interfaces, and
extensive computer capability.
Mechanical scanners provide a 7-axis
automated scanning of complex engine
part geometries; automated probe
positioning and changing to enable
inspection of multiple features; and
set-up “Standardization” reference
artifacts for automated NDE set-up
and diagnostics. e RFC system was
originally applied to the USAF F-100
engine and provided cost savings that
far exceeded the development costs.
e system provided inspection of
simple geometries (holes, radius areas,
slots, etc.) in initial applications, but has
been extended to complex geometries
| C | V. XXII | N. 23 | - | 2017 |
A T G R E C S   I  G T E
