ELEC50007 – Electromagnetism

Santos Santiago Garcia Valls

02390906
 Task 1

The relative permittivity (ϵ r ) depends on the material surrounding the charge. In a uniform
environment, ϵ r is constant and equal to 1 everywhere, leading to a graph with a uniform colour. In
this task, the EP matrix is configured to have a value of 1 across the entire area, representing a
homogeneous environment. This results in the first graph showing a uniform distribution, indicating
no variation in ϵ r .

The line charge is set to an initial potential of 10V at its centre. From this point, the potential
decreases radially outward toward the edges of the enclosure, where it reaches 0 to satisfy the
boundary conditions. This decrease follows a logarithmic relationship ϕ (r )∝ ln ⁡(r ). This logarithmic
behaviour is characteristic of a line charge in a two-dimensional system, where the potential
diminishes steadily but non-linearly as the distance increases.

The electric field lines originate from the centre, radiating outward because the positive field of the
line charge repels positive charges. The lines are denser near the centre, where the electric field is
strongest, and they spread out as the field weakens with distance. The magnitude of the electric
λ
field is inversely proportional to the distance r , described by E ( r )= . At the edges of the
2 π ϵ0r
enclosure, the electric field lines bend to remain perpendicular to the equipotential lines at ϕ=0,
ensuring the boundary conditions are satisfied.

On the horizontal centreline (x-axis), the electric field components in the vertical direction ( E y )
cancel out due to symmetry, leaving only the horizontal components ( E x ). The magnitude of E x
1
decreases with , where r is the horizontal distance from the center. At the centre ( x → 0), E x
r
becomes very large, while it diminishes to 0 as x → ∞ . This pattern is consistent for both positive
and negative directions. Similarly, on the vertical centreline (y-axis), the horizontal components ( E x)
1
cancel, leaving only the vertical components ( E y ). The magnitude of E y follows the same
r
relationship as E x , but along the y-axis.

The electric flux density ( D ) is related to the electric field ( E ) through the equation D=ϵE where
ϵ=ϵ 0 ϵ r . In this case, since ϵ r =1, the relationship simplifies to D=ϵ 0 E . This indicates that D and E
are aligned in the same direction. Additionally, because D is directly proportional to E , the graphs
of D will exhibit the same symmetry and overall behaviour as the graphs of E .
 Task 2

In both Example 1 and Example 2, the line charge is centred within

the enclosure. The primary difference between the two examples
is the shape of the boundary. In Example 1, the boundary is a
square with a side length of 100 units, held at a potential of ϕ=0
(grounded). In Example 2, the boundary is circular, defined by a
IMAX −1
radius SPAN. SPAN ¿ , with IMAX = 101. This gives the
2
circular boundary a radius of 50 units.

Data from the simulation shows that in Example 2 is twice

Ex ( x )
the magnitude of E x ( x )in Example 1. This difference arises because the circular boundary in
Example 2 is held at a potential of −10V, while the square boundary in Example 1 is grounded at
ϕ=0. As a result, the potential difference ( Δϕ ) in the circular enclosure is double that of the square
enclosure: Δ ϕ 1=−10 , Δ ϕ 2=−20 .

−Δϕ
Using the formula for the electric field in one dimension E x = , and we know that Δ ϕ 2=2 Δ ϕ 1,
Δx
leading to E x 2=2 E x 1. To create a log-log plot, only the subset E x ( x )for x >0is used since log 10 (−x )
is undefined.

Example 2 produces an electric field with double the magnitude of Example 1, which can be
attributed to two factors: a larger potential difference and better symmetry. The circular boundary
creates a more uniform and concentrated field because it is equidistant from the central line charge
at all points, maintaining a radius of 50. In contrast, the square boundary introduces corners and
edges that distort the symmetry and temper the field’s intensity.

In both examples, E x ( x )decreases as the distance from the line charge increases, following a power-
law relationship, E x ( x )=C ⋅ x n, where x is the distance from the line charge, n is the decay rate, and
C is a scaling factor determined by the potential difference ( Δϕ ). The decay rate, represented
by the exponent n , is obtained from the slope of the power-law trendline in the log-log plot. For
Example 2, n=−1.019, while for Example 1,n=−0.973. This indicates that the square boundary
causes a slower decay of the electric field compared to the circular boundary. The geometry of the
square allows the field lines to spread unevenly, reducing the efficiency of
the field’s attenuation relative to the circular setup.

Task 3

In this scenario, the electric field is not established by internal electrodes,

unlike in the previous examples. Instead, the uniform field is created by
assigning fixed potentials to the top and bottom boundaries of the
enclosure, with the top boundary set to +10V and the bottom boundary
to -10V. This configuration ensures that the field points consistently
downward, with a total potential difference ( Δϕ ) of 20V across the
enclosure. The left and right boundaries of the enclosure are set to vary
smoothly from -10V to +10V, creating a gradient across the sides that supports uniformity in the
field. The setup is implemented in the code as follows:

VF = 10;
V0(1, :) = -VF; %Bottom boundary
V0(IMAX, :) = VF; %Top boundary
V0(:, 1) = linspace(-VF,VF,IMAX); %Left edge, smooth
gradient
V0(:, IMAX) = linspace(-VF,VF,IMAX); %Right edge, smooth
gradient

To modify the setup so the left and right boundaries are grounded (set to 0V), the commands are
updated as follows:

V0(:, 1) = 0; % Left boundary grounded

V0(:, IMAX) = 0; % Right boundary grounded

This change disrupts the uniform field, as the smooth variation in potential along the sides is
replaced with a fixed value of 0V. The resulting field is no longer purely vertical. With smooth side
potentials, when the left and right boundaries have linearly varying potentials, δϕ ⁄ δx=0 ,resulting
in E x =0. The electric field ( E ) is entirely vertical, with a uniform E ycomponent due to the
consistent vertical gradient in potential.

With grounded side boundaries, the left and right boundaries create a horizontal variation in
potential, leading to E x =−δϕ ⁄ δx ≠ 0.Near the sides, the field lines bend to remain perpendicular
to the 0V equipotential lines. As a result, E x is strongest near the edges due to the horizontal
gradient. In the central region, the potential gradient remains predominantly vertical, so δϕ ⁄ δx ≈ 0 ,
and E x ≈ 0 . E y remains nearly uniform in the center but weakens near the sides as the field bends
toward the edges.

If the central region of the enclosure were made wider, the effect of the 0V boundaries on the sides
would
diminish. In
this case the
electric
field in the
centre
would
remain
primarily
vertical,
with and E ymaintaining uniformity. Distortions near the sides would have less influence on
the central region, allowing it to better approximate the desired uniform field. Note that the
potential remains uniform inside it due to the material's properties, while the cylinder concentrates
the field lines around its boundaries, creating a stronger localized field.

Task 4

At y=0, E x reaches its maximum value because the potential gradient, δϕ /δx , is greatest in this
region. This is a direct consequence of the electrodes being closely spaced, which creates a steep
change in potential over a very short distance. Using the relationship E x =−δϕ /δx , the large
potential difference (δϕ ) and small spacing (δx ) result in a significant field strength. Additionally, the
magnitude of the electric field decreases as 1/r 2, where r is the distance to the electrodes. Since r is
minimized at y=0, the field strength is highest here. When y ≠ 0, the field weakens symmetrically as
the distance to the electrodes increases. The potential gradient along the x-axis diminishes because
the field lines bend outward and become more dispersed as y moves away from the centre. This
spreading of the field lines reflects a reduction in field strength, which decreases symmetrically on
both sides of the x-axis.
Along the y-axis, E y is approximately zero because the electrodes are placed symmetrically, one at
−10 V and the other at +10 V . At each point along the y-axis, the vertical contributions from the
two electrodes cancel out. This happens because the distances from any point on the y-axis to the
two electrodes are equal, resulting in equal and opposite field contributions. Consequently, the
potential along the y-axis remains constant, leading to δϕ /δy ≈ 0, which implies E y ≈ 0.

The

equipotential lines exhibit a symmetric pattern about both the x- and y-axes due to the symmetrical
placement of the electrodes and the 0V condition imposed by the grounded enclosure. Near the
electrodes, the equipotential lines are tightly packed and nearly vertical, indicating a strong
potential gradient and thus a strong electric field. As the equipotential lines approach the enclosure
boundaries, they bend sharply to meet the 0V condition of the grounded boundaries. This bending
represents fringing effects, which are more pronounced near the corners of the enclosure. These
effects diminish toward the centre of the enclosure, where the field is dominated by the interaction
between the electrodes rather than the influence of the boundaries.

By fixing the potential at 0V along all boundaries, the enclosure confines the electric field to the
interior of the square, preventing external influences from affecting the simulation. This ensures
that the system is finite and well-defined. Near the boundaries, the grounded condition creates
fringing effects as the field lines curve to remain perpendicular to the equipotential lines. These
effects are most significant at the corners and edges, where the symmetry is slightly distorted.
However, in the central region of the enclosure, the influence of the grounded boundaries
diminishes, allowing the system to approximate the behaviour of an unbounded dipole
configuration.

Task 5

We recall the relationship E=−∇ ϕ, meaning the electric field depends on the gradient of the
potential throughout the enclosure, including regions outside the capacitor plates. To ensure a
constant electric field inside the dielectric slab, the gradient of the potential ( ∇ ϕ ) must be zero
outside the capacitor. This eliminates any interference from an external field, allowing the field
inside the slabs to remain undistorted.

In an ideal capacitor, the plates are infinitely large, and the field lines between the plates are
perfectly perpendicular. In this situation, the plates dominate the surrounding space, and the
potential gradient outside the plates is zero since no other influences are present. For a real, finite
capacitor, however, the plates are not isolated from their surroundings, leading to field distortions.
To approximate an ideal infinite capacitor, it is necessary to align the values inside and outside the
slabs, ensuring they coincide.

Along the entire vertical axis ( y ), E x =0. This is due to the horizontal symmetry of the capacitor
plates, which are treated as infinite in the horizontal direction. This symmetry eliminates any
horizontal potential gradient (δϕ /δx=0), resulting in E x =0at all vertical points.

Outside the capacitor plates, E y =0 because the ideal slabs completely confine the electric field
within the capacitor. This behavior is achieved by setting the potential gradient outside the plates
to zero, ensuring no field exists beyond the slab region. Within the plates, E y is uniform and
directed perpendicular to the plates. The relationship E y =−Δϕ/d holds true, where Δϕ is the
potential difference, and d is the separation between the plates. Both Δϕ and d are constants,
ensuring a uniform electric field inside the capacitor.

Introducing a dielectric between the plates alters the relationship between the electric field ( E )and
the electric flux density ( D ). The vertical component of the flux density is given by:
 D y =ϵ ⋅ E y, where ϵ=ϵ 0 ϵ r .

Here, the dielectric has a relative permittivity of ϵ r =4 , compared to air with ϵ r =1. As a result, the
magnitude of D inside the dielectric slab is four times greater than E . However, the variation along
the axis remains identical for both D and E .

If the boundary potentials were not defined with a linear variation, the field lines inside the
capacitor would bend toward the lateral edges. This occurs because field lines must always remain
perpendicular to the equipotential lines. Such bending would introduce edge effects, disrupting the
uniformity of the field inside the slabs. To maintain a consistent field, the linear variation in
potential is essential, as it minimizes distortions and ensures the field lines remain evenly
distributed between the plates.

Task 6

Between the plates, the equipotential lines are almost parallel and evenly spaced, indicating a
uniform electric field in this region. This behaviour closely resembles the infinite capacitor setup
seen in Task 5, as the parallel plates dominate the field between them. However, near the edges of
the plates, the equipotential lines curve outward as the field spreads out due to the finite width of
the plates. This curvature is caused by edge effects, where the electric field is distorted by the finite
geometry of the plates and the influence of the grounded enclosure. As the plates only span
0.8 × SPAN , the field lines are no longer perfectly perpendicular, reducing field uniformity near the
edges.

Outside the plates, the electric field weakens, and the equipotential lines spread further apart,
indicating a decrease in field strength. This is because the field becomes less focused, and the
influence of the plates diminishes with distance. The finite width of the plates also causes the field
lines to curve outward and lose alignment, introducing horizontal components near the edges. The
grounded enclosure walls also influence the field in this region by pulling field lines toward the
edges and forcing them to terminate orthogonally at the walls, further distorting the field near the
boundaries.

The dielectric slab increases the flux density within the plates. The relationship between the electric
field ( E ) and the flux density ( D ) is given by D=ϵ E , where ϵ=ϵ 0 ϵ r . Since the dielectric slab has
ϵ r =4 , the flux density within the slab is four times greater than in the surrounding air. The
equipotential lines are closer together within the slab, indicating stronger electric fields, while
outside the slab, the flux density and field strength decrease. This amplification of D by the
dielectric significantly enhances the capacitor's ability to store charge.

The electric field distribution shows a nearly uniform field between the plates, where the field lines
are vertical and evenly spaced, consistent with the uniform spacing of the equipotential lines.
However, at the edges of the plates, the field lines curve outward due to edge effects, breaking the
uniformity. Beyond the plates, the field lines become more dispersed and curved as the field
weakens and spreads out. Near the grounded enclosure boundaries, the field lines bend
significantly, terminating orthogonally at the walls. This bending introduces both horizontal and
vertical components to the field near the boundaries, further disrupting the uniformity of the field
in these regions.
In addition to the uniform field region, the grounded enclosure influences the overall field
distribution by confining the electric field within the enclosure and minimizing external effects. The
grounded walls ensure that the field terminates at the boundaries, creating a well-defined and
finite system. The top and bottom walls are grounded at ϕ=0V , while the plates are at ± 10V ,
introducing additional vertical components to the field outside the plates. This vertical field
component results from the field extending from the top and bottom plates to the grounded
boundaries, further contributing to the irregularities near the edges.

Task 7
The electric field distribution is shaped by the dielectric slab's non-uniform permittivity, the
geometry of the plates, and the grounded enclosure. Between the plates, the electric field is
primarily vertical and uniform. The uniformity is enhanced by the dielectric, which concentrates the
flux density within the central region. At y=0, where ϵ r =4.5 , the field transitions smoothly
between the upper and lower regions, minimizing discontinuities in the electric flux density D . The
field strength E y, however, is inversely proportional to ϵ r , making it stronger in the lower region (
ϵ r =3) and weaker in the upper region (ϵ r =6).

The behaviour of the electric flux density D is

governed by the relationship D=ϵ E , where ϵ=ϵ 0 ϵ r . At the boundary between the two dielectric
materials, D y remains continuous due to the boundary matching condition D y , lower=D y, upper.
However, E y adjusts at the boundary, creating a discontinuity to compensate for the change in ϵ r .
This ensures that the normal component of D is conserved, as required by Maxwell’s equations.

Near the edges of the plates, the field lines curve outward, disrupting the uniformity of the central
region. These edge effects are caused by the finite width of the plates and the grounded enclosure
pulling field lines toward the boundaries, where they terminate perpendicularly. The curved field
lines near the edges introduce horizontal components to the electric field, further reducing
the uniformity in these regions.
The plot of D y along the vertical centreline ( x=0 ) demonstrates that the solution satisfies the
boundary matching conditions. In the lower region ( y <0 ), where ϵ r =3, D y is higher because the
electric field E y is stronger. At y=0, where ϵ r =4.5 , D y remains continuous, though a spike may
appear due to the sudden change in ϵ r at the interface. In the upper region ( y >0 ), where ϵ r =6, D y
is lower due to the weaker electric field. Outside the plates, D y ≈ 0, consistent with the absence of
significant flux in those regions.

Overall, the central region between the plates shows the most uniform field distribution, with the
dielectric ensuring a gradual change in field strength across the slab. The grounded enclosure
confines the field within the system, ensuring a finite configuration while introducing distortions
near the edges. The continuity of D y and the adjustment of E y at the dielectric interfaces confirm
that the solution satisfies the required boundary matching conditions. This configuration effectively
demonstrates the behaviour of a finite capacitor with non-homogeneous dielectric properties.
 Task 8

The field distribution within this microstrip waveguide is shaped by the narrow strip electrode, the
continuous sheet electrode, and the dielectric slab separating them. Between the strip and sheet
electrodes, the field is highly concentrated, as most of the field lines are confined within the
dielectric region where the potential gradient is steepest. This leads to densely packed field lines
near the strip electrode, particularly along its edges. The narrow width of the strip electrode causes
a steep potential gradient directly below it, resulting in high electric field intensity in this region.

Between the strip and sheet electrodes, the field is primarily concentrated in the dielectric slab,
where the potential gradient is steepest. This creates a strong electric field localized to the
dielectric region. Near the strip electrode, the narrow width results in a high-intensity field directly
below the electrode. The field lines are densely packed, indicating a high gradient, but fringing
effects are observed at the edges of the strip electrode, where the field lines bend outward into the
surrounding region, disrupting uniformity. The continuous sheet electrode, in contrast, provides a
broad reference potential across its entire width, balancing the strong localization created by the
strip electrode.

The grounded enclosure further influences the field distribution. Near the walls of the enclosure,
the field lines terminate perpendicularly, introducing distortions in these regions. The effect is most
pronounced near the corners, where symmetry is least preserved. These distortions create
asymmetry in the field near the boundaries and reduce field uniformity.

The electric flux density, represented by D y =ϵ E y , highlights the relationship between the
dielectric properties and the field intensity. Within the dielectric slab, D y remains nearly constant
due to the uniform electric field in the central region. However, at the edges of the strip electrode,
fringing effects reduce D y as the field lines spread outward. Outside the dielectric slab, the flux
density diminishes as the field weakens, eventually approaching zero in the regions far from the
electrodes. The flux density distribution along the vertical centreline confirms that D y remains
continuous within the dielectric, reflecting the boundary matching conditions.

To improve field confinement and uniformity, the configuration is adjusted by setting the bottom
slab to 0 V and the top strip electrode to +10 V . This modification reduces the potential difference
between the electrodes, creating a smoother transition in the electric field and ensuring more
uniform confinement within the dielectric region. By lowering the potential gradient near the
edges, fringing effects are significantly reduced, and the distortion caused by the grounded
enclosure is minimized.

% Adjust potentials to improve field uniformity

V0(1:slab_bottom, :) = 0;
V0(slab_top, strip_start:strip_end) = 10;

The electric field becomes more evenly distributed within the dielectric slab, with a uniform field in
the central region and smoother transitions toward the boundaries. The field lines remain
concentrated between the strip and sheet electrodes, while the reduced potential difference
diminishes abrupt changes at the edges. This configuration allows the field to better align with the
design goals of the waveguide.
 Task 9

To approximate the total electric flux leaving the sheet electrode, the integral of D y /ϵ 0 was
calculated along the horizontal centerline. Since y=0 is the closest accessible point to the sheet
surface, the values obtained closely represent the normal component of the electric flux density.
The integration was performed numerically using Excel and the trapezoidal rule, dividing the curve
into small trapezoids with a step size of ∆ x=1 mm . Adding up these contributions gives the result:

( )
49
ϕ 10
−3
Dy D
=∑ ∙ ( x )+ y ( x−1 ) =−254.5 5
ϵ 0 x=−48 2 ϵ0 ϵ0

Dy Dy
Where (x ) is the value of at x .
ϵ0 ϵ0
Using Gauss’s Law, adapted for this configuration:

D ρ q
∯ϵ ∙ d s=∭ =
ϵ0 ϵ0
0

The chosen surface encloses the entire horizontal sheet electrode while cutting through the
dielectric slab between the electrodes. This ensures that the charge enclosed corresponds to the
total charge stored on the sheet electrode. Since the electrodes are treated as having unit length in
q
the z-direction, the result represents charge per unit length: =± 254.55 ∙ ϵ 0 , for each electrode.
l

This gives the experimental capacitance per unit length as:

( Cl ) experimental
=
q
U ∙l
=25.455 ∙ ϵ 0

To compare this with theory, an analytic estimate for a parallel plate capacitor is used, adapted to
fit the microstrip geometry. For a perfect capacitor, the capacitance per unit length is given by:

( Cl ) q A ∙ ϵr ϵ0 w ∙ϵr ϵ0
= = =
analytic U ∙l d∙l d

where d is the separation between the electrodes, A is the electrode area, and w is the width of the
strip. Since the field is mostly confined to the width of the strip, the dimensions of the strip are
used for this calculation. Substituting the values:

( Cl ) analytic
=
SPAN ∙ 0.5∙ 4
SPAN ∙ 0.1
∙ ϵ 0=20 ϵ 0

The analytic value is smaller than the experimental result. This discrepancy arises because the
analytic calculation assumes a perfect capacitor with no fringing effects, while the experimental
setup accounts for the physical width of the strip electrode and additional field contributions from
outside the main dielectric region. These fringing fields increase the effective capacitance in the
experimental configuration. This distinction is particularly important in microstrip applications, as
edge effects and nonidealities significantly impact performance.