Session 10D3

Approximation Technique For Determining Gain

And Radiation Pattern Of The Horn Antenna

Paul D. Ewing
Development Engineer
Oak Ridge National Laboratory
Oak Ridge, Tennessee

Abstract
An approximation can be used to avoid the complex
mathematics and computation methods for
calculating the gain and radiation pattern of the
electromagnetic horn antenna. Because of the
curvature of the antenna wave front, calculations
using conventional techniques typically involve
solving the Fresnel integrals and using computeraided numerical integration. With this model,
linear approximations give a reasonable estimate
of the gain and radiation pattern using simple
trigonometric functions, allowing a hand
calculator to replace the computer. The case of
the E-plane horn antenna was used to evaluate
this technique, applying selected criteria.
Results showed that the gain approximation holds
for an antenna flare angle of less than 10" for
typical antenna dimensions, and the E field
radiation pattern approximation holds until the
antenna's phase error approaches 60", both within
typical design parameters. This technique is a
useful engineering tool.

The case of the E-plane horn antenna was

used to evaluate the feasibility of the
approximation technique. The E-plane horn
antenna, shown in Fig. 1, is formed by flaring
the waveguide in the y-z plane, the E-plane of
the antenna's aperture. The horizontal dimension
of the aperture is the same as the width of the
original waveguide, and the vertical dimension is
much greater than the height of the original
waveguide.

I
Figure 1. The E-plane horn antenna.

Introduction
In general, the electromagnetic horn antenna
is a simple aperture antenna used to provide a
smooth transition for a wave traveling from a
waveguide into free space. The dimensions of the
waveguide are usually chosen to allow the
propagation of one dominant mode, and the horn
antenna will support only that mode by acting as
a filter for other modes that might be excited in
the transition, The horn antenna has the
advantages o f high gain, relatively wide
bandwidth, and easy construction. However, a
disadvantage is that the calculations for
determining gain and radiation patterns are
mathematically rigorous and involve solving the
Fresnel integral and performing numerical
integration that may require a computer. To
avoid complex mathematics and computation
methods, approximations for the gain and
radiation pattern can be used. As with other
engineering approximations,values derived from
the calculations are useful only within the
limited constraints of the approximation.
*Operated by Martin Marietta Energy Systems,
Inc., for the U . S . Department of Energy under
Contract No. DE-AC05-840R21400.

Proceedings

Gain and Radiation Patterns

Knowledge of electromagnetic horn antenna
theory has been around since the 1930s, and
derivations of the gain and radiation patterns of
the E-plane horn antenna are well
The geometry of the E-plane
documented.1' *
horn antenna in the direction of the vertical
flare is shown in Fig. 2. With the flare, waves
do not arrive at all points in the aperture at
the same time; therefore, they form a cylindrical
wave front with a phase error.
In Fig. 2, the length of the antenna flare
and the angle of the flare are represented by 1
and a, respectively. The phase error is
represented by 6 ( Y s ) , since it varies along the y
axis, and can be expressed as
6(Ys)

21 ,

(1)

where the phase constant as a function of

wavelength X is expressed as

- 1989 Southeastcon
296

CH2674-5/89/0000-0296$01.00019891EEE

and C ( p ) and S ( p ) are Fresnel integrals.

The radiation patterns represent the
radiating properties in the far field of an
antenna and are a function of the direction from
the antenna. They are the relative intensities
of the E and H fields normalized to their maximum
intensities. The E field in the E-plane of a
rectangular aperture is

For the aperture distribution given in Eq. (4),

Eq. (11) can be expressed as

Figure 2. Geometry of E-plane horn antenna.

(12)

,=

The aperture distribution for the E-plane horn

antenna is

+
-jKr

j~,

and, by substituting Eq. (1) into Eq. ( 3 ) , can be

expressed as

where Eo is the distribution across a uniformly

illuminated aperture.
Antenna gain is the ratio of the maximum
power density and the average power density in an
antenna; that is,
G

Pmax/Pave

cos2

2a fb ejK(sin@Y,

Y:/21)

dY, .

The integral in Eq. (12) has no closed form and

has to be evaluated by numerical means. Using
numerical integration, the integral will take the
form of
(13)
- Y:/21) dY, = 2b F(KbsinB,&),
/:bejK(sinBYs
where F(Kbsine,G) is the relative E field
radiation pattern for the E-plane horn antenna.
A graphical representation of F(Kbsin9,G) is
shown in Fig. 3 .

(5)

The maximum power density is

where the area of the aperture is S with a

differential area element dS, q is the intrinsic
impedance of the medium, and r is the spherical
coordinate indicating the distance to the point
of observation. The average power density is the
total radiated power for the area of a sphere,
4nr2. Therefore, the average power density can
be expressed as
IEs(Xs,Ys)1'

pave

jls

dS

0'

PhareError

00' Phaae Error

180' Phaee Error

(7)

q 4nr2

The terms in Eq. (5) can be replaced by Eqs. (6)

and ( 7 ) so that the gain becomes
kbsid

Figure 3. Relative E field radiation

pattern.

For the aperture distribution of the E-plane horn

antenna given in Eq. ( 4 ) and the phase constant
in Eq. (2), the gain is calculated to be

The H field in the E-plane of a rectangular

aperture is

By substituting the aperture distribution in

Eq. ( 4 ) , the H field can be expressed as

where

Proceedings

- 1989 Southeastcon
297

e-jKr
H E = j -

4ab

By setting the power density terms equal, E, can

be expressed as

Xr cos2 !2

Eo

-jS(p)

C(p)

cos(Kasin9)
- ;( 2 KasinQ)'

(15)

Thus, the relative H field pattern is

F(KasinQ)

cos(Kasin9)
2
1 - (- Kasin9)2

If the intrinsic impedance is replaced by

7 = 1207~= 377 for the far field and
consideration is given to the isotropic nature of
the antenna, the magnitude of the E field
strength is

(16)

Figure 4 is a graphical representation of the

relative H field radiation pattern as a function
o f the direction from the antenna.
1.C

0.1

E
*

n
E

z
0

0.1

->
-

Figure 5 shows a linear approximation of the

E-plane horn antenna wave front that will divide
the wave front and the antenna into two sections,
each acting as an independent horn antenna. With
the linear approximation, the vertical dimension
b can be replaced by
b
d = -cosa

0.4

\
\

0.2

(23)

The phase error is zero, and the angular

direction away from the antenna is approximately
the angle of the antenna flare.

0.0
I

kaslne

Figure 4. Relative H field radiation

pattern.
Gain Approximation
An approach to approximating the gain can be
the ratio of the power radiated by the E-plane
horn antenna ( P E ) to the power radiated by a
l o s s l e s s isotropic antenna (PI); that is,
G'

PI
-

'
with a linear approximation of the E-plane horn
antenna wave front being used to estimate P E .
The ratio of PI and PE can be derived by letting
the magnitude of the E field strength from the
isotropic antenna (E,) equal the magnitude of the
E field strength from the E-plane horn antenna
(E,); that is,
E'

For the isotropic antenna, the power density

(POI) can be used to relate PI and E,, since

Figure 5. Geometry O F linear approximation.

With the linear approximation, the aperture

distribution is
Es(Xs,Ys) = Eo ,
and from Eq. (11), the E field is

(24)

E,

(25)

E1

'DI

(19)

By solving the integrals in Eq. (25), the E field

can be expressed as

and

P,,

PI

E,

47rr2

Proceedings

2j

-jKr cos2 Q E, 2a
2

- 1989 Southeastcon
298

[" sin(K
d/2 sina)l(26)
K d/2 sina

Also, if the d term is replaced by its value in

Eq. ( 2 3 ) , the E field becomes
-jKr

E,

2j

Xr

cos'

;Eo

2ab

cosa

tana)
[sin(KK b/2b/2 tana

So, from the E field in Eq. (27), its magnitude

can be expressed as

IEEI

4ab

Eo cos2

tana) ( 2 8 )
;[sin(KK b/2b/2 tana
1.

However, the magnitude of E, is not

expressed in terms of PE, and the power density
o f one section of the antenna (PDE) will have to
be used to relate E, and PE ; that is,

.
.

Q rj,/
i 104

actual
approximated

0~

2-J

l
o!

--17----1--

1.0

0.6

and

1.6

2.0

2.6

5.0

5.6

4.0

4.6

6.0

6.6

8.0

Vertical Dimension blh

Figure b. Actual gain vs approximated gain-0.5, a = 5 " , b/X = 0.5 to 6.0.

a/X

By setting the power density terms equal, Eo can

be expressed as

By substituting the intrinsic impedance

q = 1207r = 377 and Eq. (31) into Eq. ( 2 8 ) ,
magnitude can be expressed as

the
* Actual

IEEI
4ab
Xr cosa

607r PE cosa

2ab

tana)
4 [sin(KK b/2b/2 tana
1'

Approximated

a l h = 0.5
a=100

With the magnitude of the E field from the

E-plane horn antenna in Eq. (32) and the
magnitude of the E field from the lossless
isotropic antenna in Eq. ( 2 2 ) , the equality set
forth in Eq. (18) can be completed and will yield
30 PI
-r

(33)

4ab

607r P, cosa

Xr cosa

2ab

a/X
cos2

tana)
;[sin(KK b/2b/2 tana
]

By squaring both sides of Eq. (33), manipulating

the terms, and substituting the value of the
phase constant K, the ratio of PI and P, and in
turn the approximated gain becomes
G'

Figure 7. Actual gain vs approximated gain-0.5, a = l o " , b/X = 0.5 to 6.0.

Radiation Pattern ADoroximation

(34)

Graphical representations of the actual gain

and derived gain approximation results are shown
in Figs. 6 and 7 where the vertical aperture
dimension b/X varies with the horizontal aperture
dimension of a/A = 0.5 and angle flares of a = 5"
and a
l o " , respectively. Using the criterion
that the approximated values fall within 3 dB of
the calculated values, the results show that the
gain approximation holds for an antenna flare
angle of less than 10".

Proceedings

A s shown in Eq. (16), the H field radiation

pattern in the E-plane is rather simple to
calculate, making it unnecessary to use an
approximation to derive it. However, the E field
radiation pattern in the E-plane is more complex
and requires the use of numerical integration for
its evaluation. From Eq. (13), the actual E
field radiation pattern can be expressed as

The pattern of the E-plane horn antenna is shown

in Fig. 8, and the relative E field radiation
pattern, as a function of direction and phase
error, has been previously shown in Fig. 3 .

1989 Southeastcon
299

f
'

The E-plane radiation pattern approximation will

combine Eqs. (36) and (37) and can be expressed
as

F,(Kbsine)

1
2

[sin(Kbsin63')
Kbsine'

sin(Kbsin0")] ( 4 0 )
Kbsine"
'

The final expression in Eq. ( 4 0 ) is actually only

half of the sum of the radiation patterns of the
two sections because of the overlap between them.

Figure 8. E-plane horn antenna pattern.

An approach to approximating the E field
radiation pattern can be similar to the one used
for the gain approximation in the sense that the
antenna will again be divided into two sections
with each acting as an independent antenna.
Figure 9 shows the geometry of each section and
its associated pattern. An estimate of the total
radiation pattern will encompass a combination of
the individual radiation patterns related by the
angles of direction Bland 9".

Graphical representations of the actual E

field radiation pattern and the approximated
radiation pattern are shown in Figs. 10 and 11
for maximum phase error values of 15"and 60",
respectively. Using the criterion that the
approximated values fall within 3 dB of the
calculated values, the E field radiation pattern
approximation holds until the antenna's phase
error approaches 60".

An approximation of the E-plane radiation

pattern will involve the radiation patterns of
the individual sections, which can be expressed
as
F(Kbsin8,)

sin (Kbsine')
Kbsine'

(36)

0.0

The radiation patterns in Eqs. (36) and (37) are

the patterns for the E-plane radiation from a
rectangular aperture. From Fig. 9 , the angles 8'
and 8" are related to the original angle of
direction e and the flare angle a of the E-plane
horn antenna by

2.0

6.0

4.0

8.0

10.0

&sin@

Figure 10. Actual vs approximated radiation

pattern--maximumphase error = 15".
Conclusions
A gain and radiation pattern approximation
can be used to avoid the complex mathematics and
computation methods typically involved in
calculating the gain and radiation pattern. A
hand-held calculator can be used to solve the
trigonometric terms in the approximations,while
a computer may be required to calculate the
Fresnel integrals in the actual gain equation and
perform numerical integration to determine the
actual radiation pattern.

and

f;

Figure 9. Geometry of E-plane radiation

pattern approximation.

Proceedings

When the flare angle of the E-plane horn

antenna is less than 10" for typical antenna
dimensions, the 3-dB criterion for the
approximation is met. However, the correlation
between the actual and approximated gain falls
off quickly when the angle flare is 10" or
greater. When the phase error is 15" or less for
the antenna, the approximation criterion for the
radiation pattern is met. Except for the regions
around the multiples of ?r, the approximated
radiation pattern holds up well until the phase
error is over 60".

1989 Southeastcon
300

References

1.0

1.

Edward A. Wolff, Antenna Analysis, Wiley,

New York, 1966.

2.

Warren L. Stutzman, and Gary A . Thiele,

Antenna Theory and Design, Wiley, New York,
1981.

3.

Richard C. Johnson and Henry Jasik, eds.,

Antenna Engineering Handbook, 2d ed, McGraw
Hill, New York, 1984.

4.

Paul D. Ewing, Gain and Electric Field

Radiation Pattern Approximations for the
E-Plane Horn Antenna, ORNL/TM-10983, Oak
Ridge National Laboratory, December 1988.

"\

'\

Actual
Approximated

0.2]

-.'4-x:6.0

'
8.0

10.0

Usin0

Fig. 4 . 6 . Actual vs approximated radiation pattern-60".

maximum phase error

Figure 11. Actual vs approximated radiation

pattern--maximumphase error = 60".
PAUL D. EWING

'PaulD. Eving has been a developpent engineer in the

Measurement and Control Engineering Section of the
ORNL Instrumentation and Controls Division since
1980. He is presently involved in the development
of electromagnetic technologies for commercial and
military applications. He has received a Bachelor
of Science degree in Electrical Engineering from
Tennessee Technological University and a Master of
Science degree in Electrical Engineering from the
University of Tennessee. Before joining ORNL, Mr.
Ewing was employed by the GTE Corp. in Cookeville,
TN and worked on the cooperative education at NASA's
Marshall Space Flight Center in Huntsville, AL.

Proceedings

- 1989 Southeastcon
301