Assessment of invehicle human exposure to electromagnetic fields

1
Assessment of in-vehicle human exposure to
electromagnetic fields

• Alastair Ruddle
• Advanced Engineering Department
• MIRA Limited

2
Overview

• Background
• human exposure
• automotive issues
• relative merits of measurements and modelling
• Results from TETRA simulations (400 MHz)
• Approximate calculations for higher frequencies
• Further work
• Summary

3
Motivation

• On-board transmitters increasingly common
• vehicle-mounted antennas
• on-board mobiles
• Greater awareness of field exposure issues
• ICNIRP guidelines
• NRPB (UK) and similar national bodies
• EU directive for occupational exposure
  (2004/40/EC)
• EU recommendation for general public
  (1999/519/EC)
• Automotive EMC directive (2004/104/EC)
• manufacturers to identify acceptable frequencies,
  powers and antenna to maintain EMC performance
• what about human exposure?

4
Requirements

• Standards specify limits for basic restrictions
• limb currents at low frequency
• specific absorption rate (SAR) at high
  frequencies
• Reference levels for uniform, non-localized
  exposures
• electric and magnetic field strengths, power
  density
• should ensure compliance with basic restrictions
• fields exceeding reference levels may still not
  breach basic restrictions, but further
  investigation is required
• reference levels not appropriate for localized
  (on body) sources
• Due to cavity resonances, in-vehicle exposure is
  also non-uniform for non-localized sources
• vehicle-mounted transmitters
• personal systems in storage spaces
• distant external sources

5
Limitations of measurements

• May not provide advance warning of possible
  problems
• physical hardware required for test
• Practical difficulties
• geometrical complexity of vehicle interior
• non-uniformity of internal fields may necessitate
  large datasets
• detailed mapping could be expensive and laborious
• Determining SAR requires additional equipment
• human phantoms
• fluid filled to mimic dielectric properties and
  allow probe insertion
• sufficiently flexible to position in vehicle

6
Advantages of simulation

• Fine spatial resolution of computed fields
• passenger compartment is resonant environment
• statistical description needed to quantify threat
• Calculation of SAR also possible if needed
• any combination of dielectric properties in
  models
• multiple occupancy effects can be investigated
• no need for physical phantoms filled with
  hazardous fluids
• No physical accessibility limitations in models
• electric and magnetic fields in every cell
• SAR in every dielectric cell
• No need for physical hardware or test facilities
• can be carried out early, when mitigation is less
  costly

7
Interior electric field distribution
8
Human simulants
CAD model allows human to be readily modified
for different vehicles, seating locations etc.
9
10 gm SAR distribution
10
Impact of driver electric field

• Empty
• vehicle
• With
• driver

11
In-car SAR models 400 MHz

• 3D numerical simulations TLM
• Car model based on CAD data for major metal parts
• bodyshell, doors, seat frames, steering gear
• Human simulants
• human shape and dimensions (based on large male)
• homogenous lossy dielectric properties
• Sources
• roof-mounted monopole
• in-car mobile transmitter (rear seat area, not in
  contact with occupants)
• Aiming to identify worst-case SAR

12
Occupied vehicle model
Spatial field output volume
13
Multiple occupancy cases
14
Occupant SAR results 400 MHz

• Mean SAR is more immediate threat than maximum
• Mean SAR limits reached at power levels producing
  average fields over empty vehicle interior above
  field reference levels for general public
  exposure
• 260 of field reference levels for roof antenna
• 195 of field reference levels for on-board
  transmitter
• Compares with 225 for uniform plane-wave
  exposure of standing human simulant in free space
  
• inhomogeneous, anatomically detailed
• Comparing average empty vehicle fields with
  existing reference levels gives similar
  protection to fee space

15
MIRA in-vehicle field exposure assessment approach
Assemble empty vehicle model
Quantify electromagnetic field exposure
Modify installation design
significantly
No
Yes
Optional limited confirmatory measurements
slightly
Augment vehicle model with simulated humans
No further action required
Estimate SAR for occupants
Action required to reduce levels
Yes
No
16
Modelling at higher frequencies

• Issues with 3D numerical simulation
• computing requirements increase very rapidly with
  frequency
• for TLM or FDTD, doubling the frequency requires
• 8 times more memory
• 16 times longer run-time
• dielectric materials may have greater impact
• Power balance calculations
• offer estimates for average field levels and
  population distributions in electrically large
  cavities
• require little detailed geometrical information
• require negligible computing resource
• expected to become increasingly reliable as the
  frequency increases

17
Approximating car windows

• Transmission at normal incidence for rectangular
  apertures
• Transmission coefficient 1 at high frequencies
• Car windows approximate electrically large from
  300 MHz

side windows rear window windscreen
18
Average internal field estimate

• For empty electrically large aperture, average
  transmission coefficient is half aperture area Ai
• For large cavity, power lost through large
  apertures is
• Equating power lost with power PR(?) entering
  cavity allows average field to be estimated

19
Average fields for 1 W CW
(Car used at 400 MHz has slightly smaller window
area, so average field estimates are higher)
20
Allowing for glass in windows

• Introduce additional aperture weighting factor
  to account for glazing in electrically large
  windows
• Coupling through dielectric layers depends on
• frequency
• polarization
• angle of incidence
• Estimate aperture weightings from transmittance
  for planar dielectric multi-layer representing
  glazing

21
Aperture weighting factors

• Assuming power is distributed equally between
  field polarizations, average transmission
  coefficient is
• For dielectric multi-layer at frequency ? and
  incidence ?, T is transmittance
• ? for fields perpendicular to plane of incidence
• for fields parallel to plane of incidence
• Weighting factors identified with

22
Field estimates with glass 1 W
solid 5 mm windscreen, all other windows 3
mm dotted all 3 mm dashed all 5 mm dash-dot
all 10 mm
23
Validation of estimates with glass
simple estimate horizontal dipole (3D
model) vertical dipole (3D model)
all windows 10 mm thick windscreen 5 mm, all
others 3 mm

Electric field at 1 W CW radiated power
24
Applying simple field estimates
Using simple power balance estimates to assess
maximum power radiated into vehicle that is
likely to comply with ICNIRP reference levels for
general public exposure
3mm windows, 5 mm windscreen all glass 1 cm
thick without glass
25
Further work

• Similar issues for other resonant environments
  that may exploit wireless-based systems
• rail vehicles, ships, aircraft, buildings
• Part of on-going collaborative project SEFERE
• Simulation of Electromagnetic Field Exposure in
  Resonant Environments
• simulation issues at higher frequencies, larger
  systems
• focused on automotive and aerospace test cases
• supported by UK government research programme
• 7 UK partners, 1 Swedish
• academic, aerospace, automotive, construction
• Further information see www.sefere.org

26
Summary

• Simulations suggest that comparing average empty
  vehicle fields with existing reference levels
  gives useful assessment of in-vehicle exposure
  threat
• assessments can be based on empty vehicle data
• Advantages of assessments based on simulations
• large datasets for non-uniform field distribution
  in vehicles
• human simulants readily included in simulations
• avoids need for hardware and expensive test
  facilities
• Simple estimates for average fields also possible
  with minimal computing and information
  requirements
• most easily for internal sources (power can be
  quantified)
• at frequencies gt1 GHz for cars (lower for bus,
  truck)