Automated eddy current Inspection of aircraft wheels

Talk originally given at Italian NDT conference, Ferrera, 1993

Joseph M. Buckley
formerly at Hocking NDT Ltd.
Abstract
The Hazards resulting from failure of an aircraft wheel during
landing, or explosion in flight, are self-evident. After a
number of accidents regular Non-Destructive Inspection of
wheels has been mandatory for many years. This is most
often done using the eddy current technique and traditionally
has been performed manually using pencil or special
purpose shaped probes.
While effective in the hands of a skilled operator it is always
difficult to confirm that such an inspection has been
performed adequately, and operator fatigue and loss of
concentration may become a problem if it is desired to
inspect more than the most critical part of the wheel. A
number of attempts have been made to automate the
process, but most have been bulky or expensive or both.
A machine is now available which can be installed easily,
requires only A.C. power and occupies less than one cubic
meter of space,
Innovative "Autotrak" software follows the wheel profile
without complex programming and allows operators with
minimal NDT experience to carry out effective and
repeatable inspections of a wide variety of wheels, checking
the full surface of the wheel in a few minutes.

Figure 1: Aircraft wheel outboard half showing major crack

Wheel failures while the aircraft is on the ground are

expensive, inconvenient and embarrassing. The results of
wheel failure in flight can be tragic. One such case occurred
in the late 1970's when the wheel of an L1011 aircraft
exploded at high-altitude. Parts of the wheel went through
the passenger cabin, causing immediate decompression and
loss of life.
The aim of wheel inspection is to prevent all such incidents
from happening.

INTRODUCTION
Wheel overview
In this paper I will describe the problems that have been
found with aircraft wheels, some of the inspection
techniques which have been used in the past, and how we at
Hocking attempted to address these problems in creating a
machine that was simple and cost -effective while meeting
the highest performance specifications.
Modern Aircraft wheels have a very exacting task to perform.
A few wheels must transmit the landing and braking forces
of an aircraft weighing typically 200 tons at 200 kph. Much
design work has been expended to create wheel assemblies
that will carry out this function thousands of times without
failure.
Normally this is achieved. However, as with any part
suffering repetitive stress, small imperfections or damage
may grow into cracks. Usually these will cause an air leak
and thus become apparent due to wheel deflation. In more
serious cases a wheel may break on landing. Figure 1.
shows a wheel that disintegrated during taxiing.

Figure 2: Exploded view of typical aircraft wheel

A typical aircraft wheel assembly consists of the following

main parts:
!
An 'outboard' wheel half containing the inflation valve.
!
An 'inboard' wheel half containing the brake assembly.
!
A tubeless tyre, inflated to a high pressure
!
Bearing assemblies, one in each half.
!
High-tensile fasteners holding the two wheel halves
together, with a rubber seal.
In typical airline use a tyre may last up to 300 landings.
When the tyre is replaced an opportunity exists to inspect
the wheel. It is vital that any inspection technique used can
locate small defects that may, during those 300 landings,
grow to a potentially dangerous size.
As tyre technology has improved, the interval between tyre
replacement has increased, meaning that smaller defects in
the wheel must be reliably located.

O:\after_s\foils\wheetalk
HOCKING NDT Ltd. 21st July 1993

Page 1

Summary of Inspection requirements

Reliability is very operator dependant.

Manual Eddy current inspection using a pencil probe.

Figure 3: Areas requiring eddy current inspection

The most highly stressed area of the wheel is the rim and
tyre seat (a) and it is here that most problems have occurred.
Most wheel manufacturers require that this be inspected at
every tyre change. When eddy current is used a typical
reference defect size is 0.75mm deep and 1.50 mm long.
The barrel (b) is also prone to some problems, particularly on
the inboard half where the brakes are attached. While not
often mandatory, many operators, particularly those with
automated machines, are carrying out inspections here also and finding cracks.
Other parts of the wheel, such as boltholes(c), the
counterbore around boltholes(d), and the ventilation holes(e)
also require inspection. This is normally carried out when the
wheel is stripped down and fully overhauled. Airline practice
and wheel manufacturer's recommendations vary, but this is
typically carried out every three to five tyre changes.
In addition, supplementary inspections such as conductivity
or hardness testing are normally carried out when there is
any indication that the wheel has overheated.
One point which should be borne in mind is that modern
aircraft wheels are expensive items and are designed to
withstand thousands of landing cycles. They will be serviced
many times during this long and arduous life, and will
inevitably sustain much minor surface damage such as
scratches, minor dents, and corrosion due to the conditions
they operate in. They will inevitably gather residues of oil, dirt
and rubber. This is perfectly normal and any inspection
technique must take it into account, In the competitive world
of air travel operators cannot afford to scrap wheels that are
perfectly serviceable.

MANUAL INSPECTION METHODS

Dye Penetrant Inspection
The wheel is stripped of paint, cleaned and coated with
penetrant. After an interval the penetrant is removed and the
wheel coated with a developer. Small traces of dye retained
by cracks in the wheel are visible against the contrasting
background of the developer.
Advantages:
!
Conceptually simple method- limited training required
!
Low capital investment - cost savings for small volume
operators.
Disadvantages
!
Very time consuming
!
Inspection is adversely affected by dirt, oil or other
contamination
!
Won't always find tight cracks- Dye may not penetrate
cracks closed by residual stress
!
Can be very difficult to reliably identify cracks in
recesses etc., for example the bead seat radius which
is the most critical area.

A standard high-frequency eddy current

crack detector, such as the HOCKING
Locator UH, is used with a pencil
probe which is scanned around the
surface of the wheel.
Various aids may be used in achieving Figure 4
a regular scan pattern. A turntable
either powered or manual helps regular rotation and allows
the operator to concentrate on scanning and observing the
instrument response. A suitable coating (such as penetrant
developer) may be applied to the wheel before scanning,
then the probe will leave a 'trail' allowing the operator to see
where he has scanned. Alternatively a suitable guide block
with several probe holes may be used to ensure regular
scanning of the bead seat area.
Advantages
!
Low capital investment - operators will probably have a
suitable eddy current instrument already.
!
Unaffected by moderate oil or dirt, removal of firm paint
is unnecessary
!
Good results if used carefully
!
An excellent backup method for investigating
indications and checking other parts of the wheel.
Disadvantages
!
Time consuming, it is usually impractical to inspect
more than the bead seat area.
!
On ameter type instrument it can be difficult to
distinguish small cracks from surface roughness or
corrosion.
!
No guarantee that all areas have been inspected Inspection quality is dependant entirely on operator skill
and integrity. Operator fatigue is a major concern.
!
No record of results.
Manual Eddy current inspection using a contoured
'bead seat' probe.
A phase plane eddy current
instrument such as the HOCKING
Phasec 1.1 is used with a specially
designed probe to scan the bead seat
area of the wheel
Advantages
!
Inspection of the entire bead
seat region in a single rotation.
!
Can distinguish between crack
Figure 5
and corrosion on the impedance
plane display.
!
Consistent scan can be easily achieved.
!
Unaffected by dirt, paint, oil, etc.
Disadvantages
!
Only inspects bead seat Region
!
Sensitivity limited (but adequate for some current
standards) as a large area of metal is inspected at
once.
!
Can be difficult to identify exact position of defect.
!
Specific probe required for each wheel type.

O:\after_s\foils\wheetalk
HOCKING NDT Ltd. 21st July 1993

Page 2

General problems of manual inspection

The Hocking Approach - WheelScan 700

!
!
!

In 1989 Hocking NDT

began development of a
new wheel inspection
system, primarily for the
RAF. (British Royal Air
Force) The intention was
to address these
disadvantages, and to
create a machine which
would be the natural
choice of all medium and
high volume aircraft
maintenance operators.

Effective inspections are very time consuming.

A high level of operator skill is required.
Operator fatigue can seriously affect the quality of
inspection
It is difficult to distinguish between small crack
indications and general surface noise and corrosion

Financially the labour dependance of manual inspection

methods means that the cost for large volume operators is a
significant factor in wishing to automate the process. From a
safety point of view, while manual inspection has found many
large defects, small flaws are likely to be overlooked. Even
after manual inspection wheel failures have still occurred.
AUTOMATED EDDY CURRENT INSPECTION
To address these problems many approaches have been
tried. Some have been very complex, using multichannel
probe arrays or specially programmed industrial robots.
Perhaps one of the more effective was that developed by
British Airways in the 1970's. Variations on this design have
been marketed by a number of companies. This employed a
pneumatic ram to lift the wheel which was then rotated at
constant speed. A probe was then applied by a gravity/spring
loaded arm and scanned across the wheel on a leadscrew.
Results were recorded on a chart recorder.
General advantages
!
Consistent scan pattern over entire flange, beadseat
and tubewell area
!
Considerable improvement in effective sensitivity.
Detection sensitivivity is the eqaul of pencil probe
inspection, but noise can be considerably improved due
to even rotation.
!
Minimal operator skill required once equipment set up.
!
Probes are Universal - work with any wheel
!
Printed (Chart) Record from each inspection.
!
Minimal operator time - extremely cost effective for
large volume use
General disadvantages
!
Relatively high initial cost discourages small operators
!
Permanently installed - difficult to move for field use.
(Normally installed in Wheel shop process) even if it
could be moved, it requires an air supply.
!
Generally single purpose; the Eddy current machine
cannot be used separately.
!
Can only be used with the wheel flange down - imposes
constraints on other processes or requires
manhandling.
!
Because it still uses an amplitude type crack detector it
can be difficult to distinguish between cracks and minor
surface denting or corrosion, particularly as:
!
The spring loaded arm tends to skip off the surface if it
is rough or corroded, Although this can be controlled by
increasing the pressure this causes increased probe
wear.
!
It can be quite difficult to find the crack after inspection.

Construction
A major requirement of the Figure 6 : Hocking Wheelscan 700
military was
transportability. It was vital that the machine could be easily
deployed to meet rapidly changing needs.
To meet this requirement several things were crucial: weight,
bulk, strength, lack of special supply requirements and ease
of setup. To achieve the necessary strength without
excessive weight we used a rugged aluminium alloy frame.
By making full use of modern electronic and mechanical
technology we were able to reduce the volume to less than
one cubic metre. By the same methods we ensured that the
only supply required is from a standard power socket 110 or
220 V 50 or 60 Hz, available in any workshop or office in the
world.
Instrumentation
To improve the performance and convenience of the eddy
current instrumentation several significant features have
been incorporated. While an impedance plane display gives
good segregation of different types of indication it requires
considerable experience to interpret. We wished to make the
instrument as simple as possible. We solved this by
providing a second 'liftoff' trace on the chart recorder,
allowing results to be easily understood: if it shows up mostly
on the main trace it's a crack, if it shows up more on the
liftoff trace it's probably corrosion or probe lift-off.
To ensure a correct test the instrument automatically sets
itself to the correct phase angle at the beginning of each test
and performs a calibration scan over a reference defect
(0.5mm slot) The calibration scan is repeated at the end of
the test. providing a permanent record of the test sensitivity

Figure 7: Typical traces from Hocking Wheelscan

O:\after_s\foils\wheetalk
HOCKING NDT Ltd. 21st July 1993

Page 3

The chart records show another major advantage of the

automated approach: if the helix is set to be smaller than the
probe field then a defect will be seen on several successive
scans, this gives a very high degree of confidence that a
defect will not be missed.

ensures that the wheel is correctly centred, presses start

and comes back 2 minutes later when the test is complete.
Few operators require such a speed, but 20 half-wheels per
hour are easily achievable for the basic inspection.
CONCLUSION

To allow the auxiliary inspections to be carried out easily up

to two additional probes can be permanently connected, for
example a bolthole probe and a general purpose pencil
probe. By pushing a button the operator selects the correct
probe and a set of preprogrammed parameters, allowing an
immediate test without spending time setting up the probe.
When the auxiliary test is completed the button is pressed to
revert to automatic operation. Thus the wheelscan is the only
eddy current machine required in the wheel shop.
Additional features include the ability to operate at
frequencies from 100 kHz to 3 MHz, and the 'FlawStop'
mode of operation where the scan stops automatically when
a flaw is indicated allowing the operator to investigate it.

As I said earlier we set out to address the technical and

practical problems of the existing technology. We feel that
we have now answered all of the technical objections for
aircraft operators contemplating automated wheel inspection.
Cost savings, reliability and safety improvements make
automated inspection the only sensible policy for larger
companies.
At Hocking we were very proud of our new machine when we
launched it, but as with any product, it's the reaction of the
customers that counts.
Three years later a customer list covering the map from
Seattle to Sydney suggests that they share our opinion!

Probe Guidance
Perhaps the most innovative feature of the wheelscan is the
'Autotrak' probe control system. We wished to allow the
probe to move freely over variations in the wheel, while
minimising any tendency to 'bounce' over rough areas. We
solved this by making the probe mechanism pressure
sensitive. At the beginning of the scan the probe moves
towards the wheel centreline until it encounters resistance. It
then moves up (or down) the wheel at a constant selectable
helix. Because the probe moves only as the driving motors
move any fast surface variations will be damped out, while
giving a minimum static pressure and thus reducing probe
wear.

Figure 8: Scan paths on Hocking Wheelscan 700

As the probe moves under processor control the orientation

is unimportant, the system can test as well flange up or
flange down. The probe will automatically follow the wheel
shape, without any additional programming.
Operation
Many otherwise excellent pieces of equipment are let down
by a poor 'user interface'. We have tried to ensure that the
task of the operator is understanding the application. The
equipment should do its job unobtrusively.
All parameters relating to the eddy current inspection are set
on the 'wheelscan E' eddy current instrument. Normally this
is just frequency, gain and alarm level. It unlikely that these
will be changed from one wheel to another.
On the wheelscan base unit the operator can set wheel lift
height (as convenient to the wheel geometry), inspection
speed and helix. For a given wheel the operator need set
only the start (flange end) and finish heights as measured
from the roller table surface,
In use the wheel is rolled onto the machine, positioned over
the spindle. Then the operator presses the lift button,
O:\after_s\foils\wheetalk
HOCKING NDT Ltd. 21st July 1993

Page 4