| ||||||||||||||||||||||||||||||||||
|
| ||||||||||||||||||||||||||||||||||
Consider a perfect radome (truncated sphere) without any panels. If we cut the radome vertically in half, we now have 2 panels and one framework seam member. As the radome protected antenna scans in azimuth, the framework shadows the antenna aperture starting from one edge, continuing over the antenna and then passing out the other side. The framework blockage length depends on the antenna scan angle and may vary from 0 up to the antenna diameter length when the framework member just bisects the reflector diameter. If we now vertically cut this radome in half once again, we now have 4 orange peel shaped radome panels. Again as the antenna scans in azimuth, the framework shadows the antenna aperture. As the antenna diameter gets larger, it is now possible to have 2 framework members pass in front of the antenna aperture ---- one framework member on one edge of the dish reflector and the other framework member on the opposite reflector edge. With actual radomes, framework shadowing is more complicated and depends on the radome geometry. Radome geometry is a term used to describe how the truncated sphere is separated into panel shapes. The mathematical process is known as tessellating the sphere. From a practical point of view, the radome must be packaged for shipment anywhere worldwide. What this means is that the maximum panel size must conform to standard shipping sizes. Taking into account shipping size constraints, two common radome geometry types are the symmetric orange peel and quasi-random geometry shown in Figures 1 and 2 respectively. Quasi-random geometry radomes may have triangular, hexagonal or pentagonal panel shapes. A geodesic radome using triangular panels is an alternate implementation of the quasi-random radome geometry.
Considering the radome geometry of Figures 1 and 2, what we see is that the framework shadow on the reflector surface is a complicated geometry problem dependent on the radome diameter, panel size, the antenna diameter and antenna scan angle. Figure 3 shows the radome framework shadows for both an orange peel and quasi-random 6.7m (22ft) diameter radome geometry protecting a 4.3m (14ft) diameter reflector antenna. To keep scattering loss low, normal practice is to use large radome panels. Therefore as the radome diameter gets smaller, the radome is manufactured from fewer panels.
| ||||||||||||||||||||||||||||||||||
On the other hand, we can purposely make the panels smaller where the framework shadowing would mimic the quasi-random nature of the radome (compare Figure 2 and 3). Such a quasi-random radome with smaller panels would have a larger transmission loss due to the longer length of framework shadow members. Figure 4 graphically shows, as a function of scan angle, the extra shadow length for the quasi-random shadow compared to the orange peel shadow. The quasi-random shadow contributes 44 percent more blockage than the symmetric orange peel geometry radome.
We wish to point out to the reader that the extra shadow length problem associated with the quasi-random geometry is a property of smaller diameter radomes where the radome panel size, antenna diameter and radome diameter have similar dimensions. For larger radomes, 8m (26ft) diameter and above, panel sizes are small relative to the antenna diameter. Here, the shadow length geometry differences becomes insignificant. We will see later that the for large radomes, the quasi-random geometry is fundamental to achieving enhanced RF performance.
When structural requirements force w to exceed RF electromagnetic specification limits, two approaches are available to reduce transmission loss and electromagnetic degradation. The first method appeals to the design shape of the framework member. The second method appeals to a process known as impedance matching where the framework electronic signature w*IFR is reduced in value. Both methods are described below.
| ||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||
A radome transmission loss (scattering loss) comparison for the symmetric orange peel and quasi-random radome is shown in Figure 7. To help with the contrast, the perpendicular joint is used with the orange peel geometry radome and the parallel lap joint with the quasi random geometry radome. Figure 7 shows that the transmission loss and framework electronic signatures w*IFR increase from 0 at low frequencies to some maximum value dependant on the detailed design and dielectric joint properties. Clearly, the framework electronic signature w*IFR has fine structure that contributes to the complex nature of the transmission loss curves. In summary, the superior transmission loss performance of the symmetric orange peel random over the quasi-random unit is both a property of the shorter shadow length and smaller electronic signature.
| ||||||||||||||||||||||||||||||||||
Clearly for a metal space frame radome, impedance matching is impossible. With a metal space frame radome, there is no approach for improving the framework electronic signature.
How the framework shadow members disturb the antenna pattern is a complex function of panel shape (framework orientation), their length and electromagnetic properties. Referring to Figure 3, the framework shadow forms line segment members. Labeling a member length li, the total shadow length l (from the transmission loss equation above) is the sum of the member lengths l = S li. Framework members each have an element pattern that scatter into the sidelobes in preferred directions according to their length, orientation and electronic signature. By keeping track of all the framework member element patterns shadowing the antenna, one has a linear array antenna. Each element of the scattering array has an element gain given by w*IFR. This linear array scattering antenna is characterized by its electric field, Es with scatter pattern proportional to |Es|2. Adding the scattering array pattern to the antenna pattern electric field, Ea, we arrive at the radome-enclosed antenna pattern, |Et|2, where Et is given by:
| ||||||||||||||||||||||||||||||||||
| Orange Peel | Quasi-random |
![]() | |
| Figure 11. Scatter array pattern comparison for orange peel and quasi-random radome geometry radome. | |
The scatter pattern figures are the standard elevation and azimuth cuts common to the antenna industry. Comparison of the scattered field over the +/- 4 degree range shows that the quasi-random scattered field has a 12 dB higher amplitude than the orange peel design. What this means is that the quasi-random radome will start degrading antenna sidelobes at the -22 dB level compared to the orange peel radome at -35 dB level. Both radomes types have higher azimuth scattered energy than elevation cuts. We will see in the next section that this focusing of scattered energy is a useful property to be exploited in radome geometry panel design. For example with a radar antenna designed to minimize ground clutter returns, the orange peel scatter pattern shows (compared to the quasi-random radome) significantly smaller effect on antenna sidelobes pointed toward the ground (elevation pattern).
In contrast to small radomes, for large radomes, where numerous framework members shadow the antenna aperture, quasi-random shadow blockage provides superior antenna performance over the symmetric radome counterpart. When the radome geometry shadow pattern is symmetric, as with large orange peel radomes, the scatter pattern equation, above, may simplified by bringing the framework element pattern, framework electronic signature and framework member length outside the summation sign:

| Quasi-random | Symmetric Orange Peel |
![]() |
![]() |
![]() |
![]() |
![]() |
|
| Figure 12. Scattering pattern for 10.7m (35ft) diameter radome with quasi-random and symmetric orange peel geometry. Note the grating lobes for the symmetric radome. Antenna is the ASR-9 radar operating in S-band. | |
29-25*log(Q )
We wish to point out that the enclosure of the antenna in a fixed panelized radome introduces a critical and novel distinction between operational system performance and performance validation by certification tests. The reason is that when the antenna is operational, both the antenna and radome are stationary with the antenna pointed at the satellite. In contrast, when the antenna is undergoing certification tests, the radome and satellite are at rest with the antenna moving to construct the sidelobe pattern. Under such circumstance, the antenna aperture blockage afforded by the radome framework continuously changes as the antenna sweeps over the sky. One therefore must make sure that both the operational antenna/radome sidelobe pattern and the antenna/radome under test both meet the intended specifications.We also note that in general there is a reduced FCC operational requirement to limit SLL outside the geosynchronous plane 32-25*log(Q ). Such a reduced requirement is measured by an elevation antenna pattern cut. This comment is of practical significance for center fed antennas. In FCC/INTELSAT applications, 4 struts that run diagonally to the outer portion of the dish typically support the feed/subreflector. The scattering by these struts, or by any long scatterer illuminated by the antenna, is in the plane transverse to the axis of the strut. The strut scattering degrades the pattern in diagonal planes. In consequence, such FCC/INTELSAT qualified antennas often do not generally comply with the SLL envelope in diagonal planes. It is of tremendous practical import that the strut scattering does not degrade the pattern in either principal plane. Such antennas comply with SLL requirements in both azimuth and elevation cuts. This allows elevation cuts to be used as validation pattern measurements in transmission using a cooperating satellite. With some limitations, the like measurement cannot completely be performed on azimuth cuts as it involves sweeping the antenna main beam along the geosynchronous arc with unacceptable illumination of nearby geosynchronous satellites. Under such circumstances, antenna symmetry may be assumed for the equivalence of both elevation and azimuth pattern cuts.
Here is where the complication begins. Introducing the radome into the antenna pattern reduces the antenna symmetry to that of the radome framework symmetry. Elevation and azimuth pattern cuts for the antenna/radome combination are now independent. Indeed, it is often the objective of the radome designer to put the scattered energy into the elevation cuts so as to minimize sidelobe error for the azimuth pattern. Even though scattered energy is very small, it has to go somewhere. The problem for the radome designer is that the antenna sidelobe energy is very small as well. Fortunately, the radome designer has two tools at his disposal. The first tool is impedance matching to reduce the framework scattering loss. The second tool is radome geometry to orient the framework members to scatter energy in appropriate directions.
Figures 13 through 16 illustrate the distinction between the measured pattern of the radome-enclosed antenna under test and the actual pattern of the radome/antenna system (the pattern on the sky). The figures are for AFC's 10.4m (31ft) diameter radome protecting a Viasat 4.5m earth station antenna. Operation is at C-band for linear polarized transponders. Figure 13 is the actual azimuth cut pattern, the illumination along the geosynchronous arc when the ESA is boresighted on the primary satellite. Figure 15 is the actual elevation cut pattern. In contradistinction, the measurement of the radome/antenna is made by rotating the antenna inside the radome and observing the pattern at a fixed far field point. Each incremental rotation of the antenna with respect to the radome modifies the set of illuminated framework joints (framework shadow). The cumulative effect is significant. Figures 14 and 16 show the measured pattern. Both measured and sky patterns sets clearly comply with the SLL envelope. Compared to the sky SLL patterns, the measured azimuth pattern has somewhat higher and less distinct sidelobes beyond -9o.
![]() |
| Figure 13. The azimuth cut pattern on the sky as radiated by the radome/antenna configuration. The pattern is fully compliant with the SLL envelope. |
![]() |
| Figure 14. The apparent azimuth cut pattern measured by rotating the antenna in the radome. The measured pattern is compliant with the SLL envelope. |
![]() |
| Figure 15. The elevation cut pattern on the sky as radiated by the radome/antenna configuration. The pattern is compliant with the SLL envelope. |
![]() |
| Figure 16. The apparent elevation cut pattern measured by rotating the antenna in the radome. The measured pattern is compliant with the SLL envelope. |
![]() |
| Figure 17. Regulatory antenna pattern requirements along with radome scatter pattern level as a function of radome transmission loss. |
The FCC, Intelsat and other regulatory agencies have yet to come to grips with a set of standards suitable for a radome-enclosed ESA. While radome design criteria directs scattered energy into the elevation plane where SLL interference is a non-issue, compliance regulations preclude such an obvious solution. Another issue relates to ESA site compliance measurements as performed for a regulatory agency and radome scattering into the far out sidelobes. For example with respect to AFC’s 10.4m (31ft) diameter radome protecting Viasat’s 4.5m ESA described in Figures 13 through 16 above, Figure 17 constructs the regulatory 29-25*log(Q ) antenna pattern envelope superimposed over the scatter pattern level. The scatter pattern levels (dBi units) are shown as a function of radome transmission loss. While ESA site compliance measurements are typically limited to scan angles +/- 10-degrees or less, Figure 17 demonstrates that the radome far out scattering pattern level actually causes non-compliance issues at pointing directions not measured during certification tests; namely 12-degrees for 0.5 dB transmission loss and 18-degrees for 0.25 dB transmission loss. On the other hand, due to the limited scan range with cooperating satellite measurements, certification tests would be found compliant. It is only when the radome transmission loss is less than 0.1 dB that strict compliance is recorded for elevation and azimuth pattern measurements.
Over standard communications bandwidth, the dilemma for radome electromagnetic design is that radome technology, through impedance matching and radome panel geometry, yielding near zero (< 0.1 dB) transmission loss is beyond the present state-of-the-art. As pointed out earlier, antenna engineers have the luxury of only 4 feed support struts oriented at 45-degrees angles. Their antenna pattern non-compliance is masked by the specific elevation and azimuth antenna pattern cut directions. In contrast, radome engineers have no such comfort. Numerous radome shadow members, oriented at various angles, preclude such an obvious deception.
![]() |
| Figure 18. Construction of AFC's 41-ft diameter radome shielding a Vertex 8.1m satellite earth station. |
The following four Adobe Acrobat PDF files illustrate radome-enclosed earth station SLL performance. The azimuth and elevation pattern measurements are for a Vertex 8.1m antenna protected by AFC's 12.6m (41ft) diameter radome as shown in Figure 18. These measurements were conducted to verify Intelsat certification for C-band satellites on circular polarized transponders.
The four pattern measurements are:
- Azimuth pattern cut for copol and xpol polarization.
Transmit RHCP at 6.339 GHz.- Azimuth pattern cut +/- 1-degree for copol and xpol polarizations.
Transmit RHCP at 6.339 GHz.- Elevation pattern cut for copol and xpol polarizations.
Transmit RHCP at 6.339 GHz.- Elevation pattern cut +/- 1-degree for copol and xpol polarizations.
Transmit RHCP at 6.339 GHz.
The above cooperating Intelsat regulatory pattern measurements show "compliance" for azimuth patterns and "non-compliant" for elevation patterns. For the elevation pattern cut, SLL compliance issues begin at the measured 8-degree scan angle where the SLL pattern is approximately –46 dB down.
AFC manufactures, markets and sells worldwide satellite dish antennas, conical horn antennas, radomes, antenna feeds, microwave and waveguide components, ultra low loss waveguide transmission line Tallguide ®, and shelters. Our customers serve the broadcast, communications, radar, weather and cable industry, defense, government, and government agencies worldwide.
A complete Internet WWW AFC site index may be found in Antennas for Communications (AFC) Home Page Document Summary List. Additional radome information is contained in AFC's Radome Capability.
Top of Page Return to AFC ProfileReturn to Radome Network Home Page.