The above images are from a Nastran mesh
file imported into CableMod. CableMod accepts
this popular mechanical physical data used
in the automotive and aerospace industries.
For the efficient simulation of large complex
cable harnesses and their ground structures,
CableMod parses the 3D Nastran file into 2D
segments, which are then analysed with the
2D cross sections of the newly created complex
harness. Also, various couplings of these
complex structures can be analysed, such as
the coupling of two arms on a y split harness,
which would be extremely complex if analysed
in 3D.
Above and on the far right, one can see the
top view of the meshed structure, with the
harness design completed within. When importing
the original Nastran structure, there are
no cable harnesses present, and one must construct
these with the built-in CAD tool provided
with CableMod.
Once
the Nastran file has been imported, in order
for the user to complete a cable harness simulation,
nodes and routes must be defined with the
chassis design. These nodes and segments are
defined in the standard CAD format.
The
next parameter to define is the specifics
about the cables within the harness. With
CableMod, the user can define many different
types of cable geometries and other physical
properties. The different types of cables
that can be defined are: single wire, twisted
pair, ribbon cable, co-axial cable, multi
wire, and self defined. Because all of the
wires are based on a material list, and the
conductivity or the permittivity can be varied
for a statistical analysis, these variations
can be also used in the actual definitions
of objects. This allows the statistical bundling
algorithm to use everything for 'real world'
simulations.
Single
Wire Definition: Physical variables include:
circle or rectangle, with adjustable material,
width, thicknesses, insulator material, diameter
of insulator (or thickness). Statistical variables
include: diameter, width, thickness, and insulator
thickness.
Twisted
Pair (ideal) definition: as in the single
wire definitions, one selects a single wire
that has been defined (as above) to serve
as a basis for the wires within the twisted
pair definition.
Ribbon
Cable Definition: as in the single wire definitions,
one selects a single wire that has been defined
(as above) to serve as a basis for the wires
within the ribbon cable definition. However,
one can vary the number of wires (up to 100);
the distance (c to c) of each of the wires
and the amount of insulating material between
the wires (the 'strength'). The distance and
the strengths can be varied for the statistical
bundling.
Coaxial
Cable Definition: as in the single wire definitions,
one selects a single wire that has been defined
(as above) to serve as a basis for the conductor.
However, one can also vary the number of these
conducting wires as well as the orbit diameter.
One can also define the insulators within
the single type wire definitions, and one
can edit or accept these. The screen can be
edited (auto or manual fit) for the outer
diameter, thickness and conducting type (ideal
ground, or signal (non-ideal ground)). Because
the conductor wires, insulators, and materials
are all based on previously entered information
that allowed variations for statistical analysis,
one can also use these parameter variations
inclusive with the co-axial variables. The
statistical analysis that is specific to the
coaxial cable is variations in the screen
outer diameter and thickness.
Multi-Wire
Definition: as in the single, ribbon, multi,
twisted pair, coaxial wire definitions, one
selects wires that have defined (as above)
to serve as a basis for the wires within the
multi-wire definition. The material for the
insulator can also be selected from a previous
definition, or one can define a new one here.
For each wire within, one selects the rotation
angle (degrees) and the position of the wire
within the bundled wire. For statistical analysis,
the insulator diameter can be varied, but
since all of the wires for the multi-wire
have been defined within all of the sub-wire
projects, all of their variations can be taken
into account.
Materials
are defined in the same way as the cables
above, but as with all of the selections within
the software, the user defines an actual project
of physical properties (cable dimensions,
materials, nodes, etc.), then extracts information
that is used in the current selection. However,
if the user does not know the specifics, but
wishes to continue with the project, they
may do so. This allows the project to be almost
completed, with the retrieval of information
for the unknown object done at a more convenient
time. This allows the efficiency of the construction
of multiple projects, without any 'stall'
time due to lack of information.
Next,
the cable and material definitions must be
defined into segments, and with each of these
segments, the background permittivity can
be varied statistically. Within these segments,
the ground structure near the bundle can also
be varied for a statistical analysis, because
the ground material definition is selected
from the project list.
Next,
sources and contacts are defined, with sources
defined either in the frequency domain or
the time domain. These are also selected from
a project list, and the user can import tabular
data, eg. from a circuit analyzer.
Finally,
one selects the type of analysis, picks the
cables in which one wishes to see results
(probes) and one can choose RadiaSim to continue
with a fields calculation (following a I/V
distribution simulation).
Types
of analysis with CableMod: I/V Analysis, Impedance
(good for line termination when estimating
post-production cable behaviour, eg. transmission
line analysis due to the random placement
of cable bundles within automobiles/aircraft),
S-parameter (reference impedance can be statistically
varied), Touchstone (reference impedance can
be statistically varied), Partial Circuit
(SPICE, Saber (distributed or lumped), with
cell length definition), and as explained
in the paragraphs above, one can have histograms
of the various types of physical and electrical
properties of the cable harness design, allowing
the user to focus in on the most probable
best, worst, and actual electrical behaviour
that will occur post-production.
Output
of CableMod to RadiaSim (not shown is the
output from a CableMod I/V analysis, with
the import to FEKO - see the paper
on this combination for detailed fields calculatons).
Following the analysis, the user can take
the results and complete a dynamic fields
calculation with RadiaSim. Note, the image
viewed in RadiaSim does not include the entire
chassis, because analysis would take far too
long to complete. However, the ground elements
included within the calculation are combined
automatically. Also, because CableMod breaks
a 3D simulation into a 2D simulation (see:
Parse3D to 2D), it can handle very complex
systems, with varying types of materials,
positions, and geometries. If a user was to
analyse these types of systems on a static
3D tool (FEM, FDTD), the solutions would of
course be more accurate, but they would not
account for the dynamic behaviour of actual
production harnesses, and these tools would
take much longer for simulations when compared
to CableMod. The computation time for a 3D
analysis would make a simulation of a complex
harness economically unfeasible.
Shown
here is the plane probes from RadiaSim (for
more information on the different types of
probes, and their manipulation, please see
the PCB EMC simulation here)
which assists the user in analysing the reflective
effects of the roof of the automobile, and
the transmission effects of the harness on
the interior compartment. Notice the color
plot of this harness, which starts at approximately
the gas filler cap on the lower right hand
side of the chassis, and continues up the
support brace to the front of the roof. The
color plots from these probes do not indicate
a high amount of field strengths - they must
be interpreted by the user using the advanced
post-porcessor and viewer SLShow. For an even
more accurate analysis, with a large GHz analysis
ability, FEKO (MoM combined with P.O. and
the U.T.D), allows the detailed field analysis
of the complex harnesses exported from CableMod.
FEKO's
wide frequency validity allows the accurate
analysis of antennaes (higher GHz) on the
largest of structures (automobiles, aircraft,
large missles, naval ships, etc.), and due
to the implementation of UTD/PO with the MoM,
economically.
CableMod allows complex analysis on it's
own, but combined with RadiaSim and FEKO,
accurate field analysis orginating from complex
harness systems, with very large ground structures
and conducting surfaces, can be analysed.
