When the discussions turn to severe aerospace applications, it’s difficult to imagine one that is more severe than a launch vehicle/spacecraft combination. Here is a microwave systems platform operating at frequencies of 2,500 MHz and higher that has a rocket engine with 4,000°F exhaust temperatures at one end, -452.5°F (+4°K) at the other end, 800g shock, 20 to 600g random vibration, and sound pressure levels at the location of the electronics packages that exceed 140 dBspl.
The acoustic energy is so intense that, without sound protection, electronic parts just disintegrate. Then there are the rapid changes in temperature and pressure: +80°F to -452.5°F and pressure from sea level to near vacuum in approximately 300 seconds. All this plus the added constraint that it costs $10,000 per launch pound to get the platform off the ground. That’s a great incentive to make things as lightweight as possible.
It’s a given that the maximum shielding effectiveness of an electronic enclosure is determined by the attenuation provided by the skin material. Even then, differences in shielding effectiveness can result from nonhomogeneous effects caused by variations in material thickness, forming/bending/welding, nonlinear material behavior at different radio frequencies, field intensities, and source locations. Many of these mechanical/material properties also directly affect the structural integrity of the enclosure.
In aerospace applications, particularly for high-performance mobile platforms such as aircraft and launch vehicle/spacecraft, enclosures often are constructed from hogged-out aluminum billets. This approach to making a box means that five of the six surfaces are made from one continuous piece that assures homogeneity and reduces the number of seams that need to be protected. That’s great for shielding, plus it allows the box to be hermetically sealed if necessary to reduce corrosion and prevent contamination.
Sealing must be capable of withstanding pressure changes. For example, a launch vehicle goes from sea level to near perfect vacuum in about 300 seconds. Unless the enclosure and fasteners are strong enough to withstand the internal pressure buildup, the enclosure will suffer permanent deformation. In the case of aircraft, the pressure variations from repeated flights not only deform the enclosure, they also serve as a pump to introduce water vapor into the enclosure and promote corrosion.
From a shielding perspective, hogging the enclosure from a solid billet assures that the walls have adequate thickness, the surfaces are flat, and the material stiffness is adequate to minimize or eliminate load deflection and all but does away with the nonlinear material behavior resulting from forming/bending/welding. And at 2,500 MHz, an aluminum box 1/8-inch thick has the potential to provide greater than 16,000-dB attenuation.
There’s no way to measure it, but based on the calculation, you could immediately conclude that this approach describes the perfect shielded enclosure. Unfortunately, from the shielding perspective, all of the variations are minor when compared to the performance degradations caused by apertures in the enclosures needed to accommodate cables, switches, displays, and maintenance panels.
After handling the aperture problem, there are several additional problem areas in aerospace applications that must be addressed. While not typically considered problems in most other applications, large changes in temperature and pressure, cavity resonance, shock and vibration, and corrosion degrade the performance of a shielded enclosure. But first, aperture attenuation needs to be addressed since aperture shielding effectiveness determines the maximum attenuation of the enclosure.
Aperture Attenuation
At the higher frequencies, an enclosure’s shielding effectiveness is dominated by absorption, as illustrated in Figure 1. A quick calculation for absorption losses of aluminum (µ = 1, ? = 0.64, t = 125) shows an attenuation level of
Absorption Loss(AdB)
= 3.34 t (? ?F) 0.5
= 3.34 × 125 (1 × 0.64 × 2,500)0.5
= 16,700 dB
Although the calculation is for aluminum, these levels of attenuation are typical for solid metal shielded enclosures. In most metal enclosures, it’s not the material that limits the shielding effectiveness, it is the apertures. The shielding effectiveness of apertures, and ultimately of the enclosure itself, is a function of their geometry; that is, number, area, and longest dimension. Since the worst-case parameter is the aperture’s longest dimension, this makes the longest enclosure seams the greatest offenders.
In most aerospace enclosures, the apertures that cannot be eliminated generally take the form of slots between the box and covers. The parallel slots on each side of the cover can be modeled as a pair of inefficient slot antennas. Accordingly, two apertures make up a phased array. The other two sides are at 90 degrees and do not add. Resonance effects related to the pair are determined by the spacing.
Considering the slot shielding effectiveness as the inverse of the radiation efficiency of a slot antenna does result in some error, but it is adequate for design because the shielding effectiveness is greater than the model indicates. Shielding effectiveness of a single aperture with slot opening length (L) is given by the first part of the following equation. The second part of the equation corrects for multiple equal length (L) apertures.
Slot Shielding Effectiveness(SEdB)
= 20 log [(?/2)/L] – 20 log (n)k
where: L = length of slot (meters) and L > w and L >> t
? = wavelength in meters
n = number of apertures with in ?/2
k =1 for farfield, 0.5 for nearfield
Farfield illuminated apertures are independent when the spacing between them exceeds the longest dimension such as L. For this case, the attenuation is reduced by 20 log N.
In the nearfield, the attenuation is higher because of phase and gradient differences. The term, n, related to holes located within ?/2 is an attempt to account for the nearfield/farfield effects, and although it works for estimating the value, it still doesn’t do a very good job.
When the slot length is equal to ?/2, the model assumes the aperture becomes transparent and the shielding effectiveness is equal to and remains 0 dB. This is not true at the higher frequencies where the differences in phase result in oscillatory behavior, but Murphy’s Law prevails.
Aperture Sealing
Any isolated panel/skin that is not adequately grounded/bonded to the enclosure can behave as an antenna structure. Grounding the panel at only one point will reduce the panel’s antenna efficiency, and it may even prevent it from acting as an antenna, but it will not eliminate leakage through the rest of the seam.
Since the covers must be removable for installation, maintenance, and repair, closely spaced screws or clamps must be used. The spacing depends on the RF frequency. For example, since FMHz × ?m = 300 m/µs, at a frequency of 2,500 MHz,?/2 is 2.36 inches.
This is why most EMC design guides recommend cover screw spacing of 1 to 1.5 inches. It is necessary to have the screw spacing less than ?/2. Most texts recommend less than ?/20. Based on laboratory data, ?/50 would even be better.
The best all-around seam configuration would be made like the lid of a paint can. It’s capable of attenuations greater than 120 dB, but its reusability is limited.
With aerospace applications, the two primary seam configurations used are the compression seam and the shear or wiping seam. Fastener spacing can be increased in compression-seam applications by using RF gasket material between the mating surfaces. Fasteners used for maintaining contact between the mating surfaces can be eliminated altogether if the seam design configuration and RF gasket can work together in shear.
The two RF gasket materials most popular for aerospace applications are beryllium copper (BeCu) spring fingers and silver-loaded elastomers with the BeCu flat spiral wound spring and Monel wire mesh configurations vying for third place. The BeCu spring products have the greatest shielding effectiveness of >20 dB over the widest frequency range combined with low compression force and the capability of being used in shear.
Spring fingers have the lowest compression forces of all the gaskets used in aerospace applications. Silver elastomers have very high shielding effectiveness of 100 to 120 dB, and Monel is not far behind with very good shielding effectiveness of 80 to 100 dB. Except for special applications, any of these materials will provide the necessary attenuation. The choice generally is based on the mechanical and corrosion properties.
Seam Configurations
The popular aerospace seams are illustrated in Figure 2. The compression seam tends to be the most frequently used type of seam, especially for existing enclosures being retrofitted as shielded enclosures. This fixed configuration is a permanent solution that will not be repeatedly opened and closed.
In this application, panels overlap the perimeter of the apertures and can be sealed with solid metal, wire mesh, or elastomer gasket materials. Since the gasket-material compression forces are normal to the panel and the attenuation is a function of compression force, the cover panel fasteners or clamps must be adjusted to uniformly compress the gasket material around the perimeter.
Some gasket materials are more sensitive to variations in compression force than others. For example, BeCu spring fingers have 6 dB to 10 dB change in attenuation from minimum to maximum compression whereas metal-filled elastomer materials may have 60 dB or more. This wide variation can be handled by torqueing the fasteners to the same value or constructing the box or panel with built-in positive stops. Positive stops also are important in reducing vibration problems.
Because the compression forces required differ for each type of gasket material due to their inherent spring force, the screw spacing must be adjusted accordingly. For example, BeCu spring fingers have compression forces that run 5 to 20 pounds per linear foot; solid elastomer materials, which essentially are incompressible, have compression forces that run 50 to 100 pounds per square inch. Using elastomers requires stronger enclosures and fasteners, and the fasteners need to be located much closer together.
The shear seam is a dynamic configuration especially suited for covers and access panels that will be repeatedly opened and closed and must be treated differently than the compression seam. This type of joint is constructed in several configurations including pan-edge, knife-edge, and modified knife-edge. These designs align the RF gasket’s mechanical forces parallel to the panel surface and can eliminate the need for fasteners to preserve shielding effectiveness.
Fasteners still may be required for shock, vibration, atmospheric pressure changes, and corrosion control. Metal-finger gasket materials typically are used for this application, but BeCu flat spiral coils can be used based on the seam design configuration.
Besides having the highest shielding effectiveness over the widest frequency range, shear seams are self-cleaning. In addition, they are resistant to environmental factors such as temperature, corrosion, and solvents, and they don’t creep.
Because of the difficulty in retrofitting shielding to a shear seam, this design configuration should be considered from the beginning of the project. For applications that also have environmental requirements, the solid metal gaskets are paired with an environmental seal. This permits the equipment to use gaskets that are optimized for the specific requirement, virtually eliminates galvanic corrosion problems, and provides the best overall performance at the lowest life cycle cost.
What has been described is an outstanding shielded enclosure in a benign environment. Unfortunately, most aerospace environments are not benign.
Shock and Vibration
If the shielded enclosure is considered a rigid body and mounted on a platform subjected to shock and vibration, then the enclosure will be displaced by the same amount as the platform. Regrettably, the enclosure is not a rigid body. All RF gaskets are springs.
Some gaskets, like BeCu spring fingers, have linear spring constants while others, like metal-filled elastomers, are nonlinear. The combination of the mounted enclosure, the cover, and the RF gasket material represents the mount, the suspended mass, and the spring of a mass-spring-damper system, respectively. Such a system will respond to a displacement forcing function differently at its resonant frequencies than at other frequencies. A comparison of three gasket systems is shown in Figure 3.
If the system is at rest and the cover is displaced from its resting position and then released, it will oscillate at its natural resonant frequency, which is given by
F =1/(2?) (K/M)0.5
where: K = spring stiffness (N/m)
M = cover mass (kg)
When the vibration driving frequency is at or below F/2, the transmissibility factor is approximately 1.0, and the cover displacement will be nearly the same as the mount displacement. When the driving frequency increases to 1.4142F, the transmissibility factor is back to 1.0. When the frequency is well above the resonant frequency of ~3F or more, the transmissibility is reduced greatly.
But at the resonant frequency, the transmissibility factor for an undamped system can be multiplied by 20 or more. The cover can violently depart from the enclosure, and if it does, you wouldn’t want to be in its path. That’s not likely to happen with the small enclosures that are being discussed.
The real problem is the loss of shielding effectiveness that results because the RF gaskets are not able to maintain contact with the wildly accelerating cover. This is more of a problem with the elastomer materials because of their mechanical hysteresis. Internal frictional forces within the material limit the rate at which they can change their configuration.
There are several approaches for handling the shock and vibration problem:
• Use positive stops and tighten the cover against the stops. If the cover can’t move relative to the enclosure, the vibration problem disappears.
• Add damping. This can be in the form of an elastomer gasket and be either an environmental seal or an RF gasket. This is especially important when using solid metal gaskets. Metal springs of virtually any shape provide almost no damping, and even though they can follow the movements of the mating surfaces, their transmissibility at resonance can be very high. They benefit greatly from the addition of damping materials.
• If the vibration frequencies of the platform are known, it’s possible to mechanically detune the enclosure by adjusting the cover mass and selecting RF gaskets so that the enclosure resonant frequencies do not correspond with the driving frequency.
• Be careful when using elastomers. Their electrical characteristics such as volume resistivity vary depending on whether the material is used under static or dynamic conditions, the amount of deformation, the frequency of the driving force, and the operating temperature.
• Utilize shock mounts. Just make certain that there is not a driving force coupling problem between the shock mount and cover.
Corrosion
The other major environmental problem is corrosion. Although there are many forms of corrosion, the two primary areas of concern with regard to aerospace equipment are fretting corrosion and electrolytic/galvanic corrosion. Although the mechanisms that produce these two types of corrosion are significantly different, they both result in very similar EMC shielding problems.
During the corrosion process, there is a loss of material and structural integrity that often results in the creation of holes in the shield, causes failure of joints and bonds, and reduces the compression forces required to adequately compress RF gaskets. The corrosion process forces contact surfaces apart and deposits insulating oxides or other nonconductive metallic salts between the surfaces that create high RF impedance joints. As a minimum, these high impedance contacts result in a loss of RF integrity.
In many cases, the problem is worse because the corrosion byproducts result in the development of nonlinear junctions. In the simplest case, the nonlinear junctions produce new, unwanted signals. For two signals—which do not need to be within the enclosure—mixing creates the sum and the difference as well as the originals for a total of four signals.
Fretting corrosion is closely related to shock and vibration. It occurs whenever two materials rub together in the presence of an electrolyte. It could be described as a combined mechanical-chemical process. Aluminum is particularly susceptible to this type of corrosion although it can affect any material that depends on its oxides for protection.
In fretting, the protective oxides are rubbed away which exposes unprotected material that oxidizes and then gets rubbed away again, and the cycle continues. This eventually wears away the surface and simultaneously puts a nonconductive oxide layer between the RF gasket material and the mating surface. Without a good electrical bond across the joint, an RF leak occurs, and the shielding effectiveness is degraded.
Electrolytic and galvanic corrosion are both electrochemical processes. They differ only slightly in the way the process occurs and the material corrodes. In both cases, there are three conditions necessary for corrosion to occur:
1. There must be an electrochemical potential between the contact surfaces. The potential differences can be created by dissimilar materials with different free energy levels (galvanic corrosion) or by external DC currents in structure (electrolytic corrosion).
Of the two, galvanic corrosion generally is the worst EMC problem because of the dissimilar metals that are used. The rate of corrosion depends upon the amount of current and the characteristics of the electrolyte. Since the current for galvanic corrosion is determined by the material dissimilarity, materials must be chosen to minimize dissimilarity.
Figures 4 and 5 can be used together to provide guidelines on what materials can be paired and the protection required. Unfortunately, even if the contact materials are the same, corrosion will occur although more slowly. There always are localized dissimilarities at grain boundaries, stress regions, impurities, dirt, and differences in electrolyte concentrations across the surfaces.
2. The contact materials must share an electrolyte. This is an ionic conductor, usually a weak acid or alkaline solution. It is nearly impossible to eliminate because temperature variations that cool the equipment below the ambient dew point result in condensation of water vapor on the cooler surfaces. Even so, the addition of sealants and environmental gaskets will reduce electrolyte penetration and slow the rate of corrosion.
3. There must be an electrical conducting path between the two surfaces. This is absolutely required to sustain the electrochemical reaction. If there is no current path, the reaction will stop because equilibrium is reached in the reaction. Having a nonconductive surface would stop the corrosion, but it is impossible to eliminate the conducting path in the shielding layer.
There are some other important points regarding corrosion:
• Since the electrolyte is on the surface of the material, it is the material’s surface that determines the corrosion rate. As a result, plating or other surface treatments can be used to protect the material. For aluminum, the choice often is irridite or alodine. Both of these coatings are conductive. Anodizing should not be used because it is an insulator.
• Carbon and carbon-based materials essentially are dissimilar to any other material and will promote the corrosion of the contact material.
• If it is necessary to use two materials that are widely separated on the electrogalvanic series, protect both of them to prevent their contact with electrolyte or use an interstitial material(s) between them that is compatible with both.
• The environment where the equipment will spend a lot of time determines the maximum potential differences between galvanic couples. If the equipment will be in a harsh/exposed environment, the electrogalvanic potential between dissimilar materials should be less than 0.15 V. For a sheltered but uncontrolled environment, the dissimilarity can be up to 0.25 V or less. In a protected controlled environment, the dissimilarity can be up to 0.5 V or less; with good temperature/humidity control, it could be as much as 0.6 to 0.65 V. Keep in mind that warehousing and transportation typically are not controlled environments.
Summary
An RF shielded enclosure is not a mechanical solution to an electrical problem. It is an electrical solution to an electrical problem. However, it still has to function properly in its operational environment. That may require a lot of mechanical design expertise.
Ideally, the enclosure should be constructed with no seams. Maybe someday a high-strength conductive paste will allow enclosure manufacturers to approximate the no-seam configuration. Meanwhile, most enclosure manufacturers use RF gasket materials to maintain contact across the seams.
Depending on the seam design used, gaskets will eliminate the need for fasteners or at least permit them to be spaced farther apart. The various RF gasketing/shielding materials have substantial differences in their electrical and mechanical properties. These differences must be carefully considered based on the application.
If it is an aerospace application, the RF gasket’s relationship to the shock, vibration, and corrosion problems must be evaluated. Some RF gaskets are better suited for some applications than others.
About the Author
Ron Brewer currently is a senior EMC/RF engineering analyst with Analex at the NASA Kennedy Space Center. The NARTE-certified EMC/ESD engineer has worked full-time in the EMC field for more than 30 years. Mr. Brewer was named Distinguished Lecturer by the IEEE EMC Society and has taught more than 385 EMC technical short courses in 29 countries and published numerous papers on EMC/ESD and shielding design. He completed undergraduate and graduate work in engineering science and physics at the University of Michigan. e-mail: [email protected]
March 2009