The complex (balun included) half wavelength dipole

Geometry (Problem) Formulation

antenna is placed parallel to the length side of the model

and is oriented along the y-axis as shown in figures 6 &

In our study, a numerical model of the complex half

7. Our dipole and electrical geometry is meshed using a

wavelength dipole antenna is placed near the numerical

cubic cell size of 0.3mm and is surrounded with a LIAO

phantom (APREL Laboratories Universal Phantom)

absorbing boundary with 20 cells of separation from all

figure 7 and is filled with a simulation liquid meeting

geometry facets. This expands the total electrical

the permitivity and conductivity requirements for the

geometry  solution  space  to  a  dimension  of

applicable frequencies. Application of the FDTD

(115.5x62.7x112.5) FDTD cells of interest (electrical

method requires determination of spatial and temporal

geometry) total 2.7 million. These calculations were

aspects before commencing the calculation where the

performed using a Pentium Dual-XEON workstation

cell size should be λ/10 or less at the highest frequency

and Remcom XFDTD software. The electrical geometry

of interest. For validation calculations, λ/20 or smaller

(including the complete Universal Phantom figure 7),

cells are appropriate where the minimum cell size for

cell size and separation distance remain fixed and do

6.0 GHz in a medium with a relative permitivity of 40

not change for all calculations presented. The only

should be;

degrees of freedom permitted for changes to the model

are with respect to the stimulus frequency 5.2GHz and

5.8GHz, and the tissue simulant parameters.

The FDTD method is applied to determine the electric

and magnetic fields calculated inside the phantom

Universal Phantom Model

together with the electrical characteristics of the dipole

antenna and feed point. The dipole remains at a fixed

distance of S=10mm between the dipole radial center

and the tissue equivalent liquid of the model. The

phantom shell is made from a low relative permitivity

and conductivity material (εr = 3.7, σ = 0.008 S/m) and

is T=2mm thick. The interior of the phantom is filled

with a tissue equivalent liquid to a depth of 100mm with

frequency dependant dielectric properties for the

frequencies 5.2GHz (εr = 36.0, σ = 4.7 S/m) and

5.8GHz (εr = 35.3, σ = 5.3 S/m). The electrical

geometric dimensions using 0.3mm size FDTD cell is

(100 x 50 x 50)mm as presented in Figure 6. Although

the phantom dimensions are significantly larger, to

figure.7

simplify the problem areas outside of the electrical

geometry are excluded.

The feed-point is excited with both Gaussian and

Problem Geometry

sinusoidal waveforms of more than 20 cycles. This

equates to 8000dt time-steps and allows courant

stability to ensure steady state is reached, for proper

calculation. The duration of the input signal has to be

chosen so as to give steady state the number of steps

necessary for the wave to propagate throughout the

whole of the electrical geometry. The bandwidth of this

source waveform is small and normally poses no

problems to grid step and distance from boundaries. The

SAR  (W/kg)  can  be  determined  (measured  or

calculated) at any point from the electric field within the

electrical geometry. Where E is the electric field in

(V/m), σ is the conductivity (S/m), and ρ is its mass

density (1.0 kg/m3) of the tissue in which the

measurement is made. The calculated results of the

dipole electrical characteristic parameters and the

calculated SAR values are normalized to 1watt of input

power and this shall be discussed in a future paper.

figure.6