Article published on juin 2008Automotive wire bundles
Fault detection accurate to within 10 cm
How can a physical defect in an automotive wire several meters long be accurately located? At present, auto repair mechanics devote considerable time to the diagnosis of wiring faults. In the future, using the diagnosis tool developed by CEA LIST researchers within the scope of the Smart Electronics Embedded Diagnosis Systems project (SEEDS), they could locate the fault in a few seconds and with an accuracy of within 10 centimetres.
Wiring failure, caution! At
present, when auto repair mechanics
are confronted with a wiring failure,
they use diagnosis tools to identify
the defective wire within a network
of up to several kilometres of wires.
But if this wire is 10 meters long
and runs through the entire vehicle
via complex paths, there is no
indication as to the location of the
point requiring repair. An the search
can take hours!
The first studies conducted on the
subject by CEA LIST researchers
focused on nuclear applications.
In recent years, the scope of these
studies has been extended to
include transport applications:
upstream research on detection
method s, sig nal f requenc y
reduction, prototyping, validation
tests on wire bundle portions, etc.
The main strength of this technology
is that it can achieve an accuracy
to within 7 cm at
present (in the
laboratory), and
probably to within
3 or 4 cm in the
future. It is based
on reflectometry,
a technique
consisting of
transmitting a
high-frequency
probe signal (100
MHz) on the wire and measuring the
alteration of the return signal, which
indicates the nature of the fault
(short circuit, broken wire, defective
connector) and its location.
'Reflectometry is a complex
technique requiring collaboration
with an advanced laboratory
to develop innovative tools',
explains Charles-Henri Garih*,
Product manager at Delphi,
one of the partners of the SEEDS
project. It perfectly complements our existing diagnosis tools,
which is why we are actively
collaborating with CEA teams.
The end objective would be to equip
every vehicle with an embedded
diagnosis chip', explains Fabrice
Auzanneau (CEA LIST). But costs
will need to be reduced to levels
compatible with the automotive
industry, which will require a lot
more work'.
A diagnosis tool for use by auto
repair mechanic s therefore constitutes an intermediate stage,
but also a market in itself.
Delphi recently received a
pre-prototype designed by CEA LIST
researchers. A series of tests and a
market analysis will be conducted
before possible industrialisation.
'We first wish to validate the tool's
suitability for the target market and
thereby ensure that it meets the
requirements of our users', explains
Charles-Henri Garih.
These systems have other potential
applications. For example, their
technology could be used to locate
wiring faults in commercial aircraft
(total of 400 km of cables in certain
aircraft!) or cruise ships (2500 km
of cables in the Queen Mary II), or
even among hundreds of thousands
of kilometres of buried electrical
distribution cables. Location
accuracy would be degraded
(approximately 10 m for electrical
cables), but it would still represent
significant progress with respect
to the tedious searches currently
required.
* Charles-Henri Garih is Head of the Automotive
diagnosis tools product line at Delphi.
