Rad Hard

Total Dose

103-104 rads

lOS-IO6 rads

Dose-Rate Upset

106-106 rads (Sl)/s

>109 rads (SI)/s

Dose-Rate-Induced Latchup

107-109 rads (Siys

>10« rads (Sl)/s



1014-1015 n/cm2

Single-Event Upset (SEU)

10-3-1 Or7 errors/bit-day

10~®-1 CH 0 errors/bit-day

Single-Event Latchup/Slngle-Event Burnout (SEL/SEB)

< 20 MeV-cm2/mg (LET)

37-80 MeV-cm2/mg (LET)

. COTS characteristics may vary unpredictably from lot to lot and even within a lot.

. Higher margins and more testing (screening) are required with COTS usage, which wffl offset lower piece part costs. • LET is linear Energy Transfer threshold.

. COTS characteristics may vary unpredictably from lot to lot and even within a lot.

. Higher margins and more testing (screening) are required with COTS usage, which wffl offset lower piece part costs. • LET is linear Energy Transfer threshold.

Whether designing satellite electronics with RAD Hard or COTS parts, a Radiation Hardness Assurance Control Plan (RHACP) is necessary to specify radiation design requirements, parts derating methods, required design margins, parts testing requirements and the process for controlling all activities related to radiation hardness. Implementation of the RHACP will help ensure the success of the hardness design and hardness verification process. The hardware is normally hardness qualified at an appropriate level, either piece part, unit, subsystem or system, whichever is economically and technically correct

Displacement fluence is any electromagnetic or particulate radiation which displaces atoms from their normal lattice positions. For nuclear weapons, neutron fluence is the primary cause of displacement In the natural environment, electrons and protons are the principal contributors. The displaced atoms and their vacancies will react with the bulk material and form stable defects in the lattice structure. These defects significantly change the equilibrium-carrier concentration and minority-carrier lifetime. In silicon solar cells, these changes degrade power output In other solid-state electronic devices, they reduce gain and increase forward voltage drop and reverse leakage currents.

We cannot harden to the neutron displacement fluence from a nuclear burst by shielding because the uncharged neutron is very penetrating and large amounts of shielding would be needed. In general, we harden to the neutron fluence by selecting devices that resist degradation by neutrons.

To protect against displacement by electrons or protons, we must shield the solid-state devices. Solar cells are shielded by a layer of fused silica, varying in thickness with the amount of shielding required. At very high ionization dose, the cover glass material darkens, reducing the solar array's power output For solid-state devices contained inside aluminum boxes, we choose the thickness of the aluminum to stop the electrons and ignore the protons, which penetrate much less in most commonly used orbits.

Delayed beta radiation flux can also be shielded effectively in the same manner as total dose, since it is composed of electrons. In contrast, delayed gamma flux cannot be shielded easily due to its high energy content (up to about 12 MeV). As an example, a factor of 10 reduction requires about 1.1 inches (2.8 cm) of high Z material like tungsten or tantalum. Mitigation of gamma debris noise spikes in sensor systems will require heavy shielding and/or pulse suppression signal processing (such as time delay integration), or even complementary satellite tasking. Even then, the gamma noise can still be high enough to cause sensor outages lasting from seconds to minutes, depending on specific sensor performance characteristics and design. For example, a fairly robust sensor with an operational capability (noise threshold plus signal to noise ratio) of 10® photons/cm2 /s will be "blind" for about 34 sec, given a 1 megaton burst at 100 km away from the satellite, using Eq. (8-12).

EMP is typically in the MHz range. At satellite altitudes, EMP intensities of a few V/m can easily cause damage and upset in unhardened satellites. To prevent this, Faraday shields can keep the radiation from entering the satellite cavities. We can also use good external grounding, interconnect all conducting parts and surfaces, employ surge arrestors, and eliminate sensitive components. In addition, designing for electromagnetic compatibility, such as shielding of cables and harnesses, will reduce or eliminate much of the potential for EMP damage. Computers are particularly sensitive to EMP, as are the following components (in order of decreasing sensitivity): semiconductor diodes in microwave applications, field-effect transistors, RF transistors, silicon-controlled rectifiers, audio transistors and semiconductor diodes in power rectifier applications.

SGEMP occurs when the incident flux of photons, both X-ray and gamma ray, creates a flux of electrons inside the satellite. Some of the energetic electrons are not stopped in solid material but emerge into satellite cavities, causing currents and fields within these cavities. At representative satellite fluence levels, these electrons can generate cable injection currents of 10-100 amperes/meter of cable length and peak cavity electric fields of several hundred kilovolts/meter. The fields then couple electroinag-netic energy into cables and other conductive elements in the cavity, and the sharp pulse of energy transmitted to sensitive components can make them fail.

SGEMP hardening uses the same methods as EMP hardening except for external shielding, because SGEMP generates inside the satellite. We can also treat internal surfaces with low-Z (atomic number) paints to reduce electron emission into cavities. Using specially designed low-response cables will also reduce SGEMP effects. Finally, we can protect input/output circuits and terminals with various devices— zener diodes, low-pass filters, and bandpass filters—to limit current or to clamp voltage.

The natural space phenomena causing single event upsets (SEUs) and other single-event effects (SEEs), as well as methods for predicting upset rates, were addressed in Sec. 8.1.5. Because shielding is ineffective in reducing SEEs, satellite systems must be designed to mitigate these effects, given that they will occur. Table 8-8 lists some classical approaches used in modem space system design. The extent to which these approaches are applied depends on the mission criticality, system upset specifications (allowable rates and outage times), and orbital environment expected. However, as indicated in the table, selection of acceptable parts is perhaps the single most important of all approaches for SEE mitigation, albeit not sufficient by itself. Not indicated in Table 8-8 is the effect of orbital altitude. While geosynchronous altitude is the worst case for SEEs (due to galactic cosmic rays), orbits that traverse the proton belt (elliptical orbits and those between about 1,200 km and 8,(MX) km altitude) will have SEEs from high energy protons, in addition to galactic cosmic rays, and the proton SEUs can be 10 times worse.

TABLE 8-8. Single-Event Effects. The effects caused by single events can be reduced by better parts, improved shielding, and process redundancy.



1. Parts Selection:

• Derate power MOSFET to 30-40% of VDS

• Vqs is rated drain to source voltage

2. Use parity and SECDED

Single error correction, double error " detection

3. Use dual or redundant logic for critical functions

2 correct outputs for decision making

4. Use watchdog timers and triple modular redundancy (TMR) In spacecraft control processor.

2 out of 3 voting logic used; switching to spare processor after repeated timeouts

5. Periodic refreshing of critical memories

Periodic switchover to refreshed memory bank

6. Use of hard latches

Eliminate soft error responses

7. Design digital circuits immune to analog circuit spikes

Long response time compared to spike transient

8. Eliminate nonrecoverable system modes and failures that could result from a soft error (bit flip)

Good design practice always required to ensure no damage and recoverable modes

Note: LET is linear energy transfer threshold; SEGR/SEB is stngle-event gate rupture/single^vent burnout

Note: LET is linear energy transfer threshold; SEGR/SEB is stngle-event gate rupture/single^vent burnout

Surface charging and resultant electrostatic discharge (ESD) due to space plasmas were addressed briefly in Sec 8.1.4, including basic design guidelines for satellite survivability. Satellites that are highly exposed to electrons (those at high altitudes, geosynchronous and highly elliptical orbits) must also be designed to survive bulk charging, in which electrons embedded in bulk dielectrics (cable dielectrics and circuit boards) and isolated conductors (such as ungrounded circuit board metallizations and spot shields on parts) build up potentials sufficient to cause discharges. Such discharges can result in anomalous upset and/or damage to electronics, much like SEUs, discussed in the preceding paragraph.

Much of the work on bulk charging is summarized by Vampola [1996], based on CRRES flight data. Mitigation approaches are indicated in Table 8-9. Designers can eliminate most bulk charging concerns simply by providing sufficient shielding to reduce both maximum current on circuit boards to less than 0.1 x 10~12 amps/cm2 and maximum total integrated fluence to less than 3 x 109 electrons/cm2 on ungrounded localized spot shields [Frederickson et al., 1992]. For geosynchronous satellites, this shielding is about 0305 cm of total equivalent aluminum (which is typically provided for total dose protection).

TABLE 8-9. Bulk Charging Mitigation Approaches. Careful planning can produce adequate solutions without large investments of time and money.




Use leaky dielectrics and bleed-off paths with < 109 ohms resistance to ground (at least 2 ground paths for contiguous areas >64.5 cm2)

Double shielded wire harness and cables

Adequate shielding (-0.305 cm aluminum) of circuit boards and part shields (vs. grounding of all metallizations and local part shields).

Signal Response Conditioning

Design circuits to be unresponsive to the relatively short, low level spurious ESD pulses which are typically less than 100 ns.

Circuit Hardness

Circuits should be designed for no damage by ESD pulses with energy levels up to 10 microjoules.

8.2.4 Strategic for Achieving Survivability

As described in Sec. 8.2.3 and summarized in Table 8-10, hardening is the single most effective survivability option. Table 8-11 presents other strategies for enhancing survivability. We use redundant nodes, also called proliferation or multiple satellites, to overlap satellite coverages. Thus, if one satellite fails, others will perform at least a part of the total mission. An attacker must use multiple attacks to defeat the space system—a costly and therefore more difficult approach for the enemy. The development of the so-called lightsat technology—light, inexpensive satellites performing limited functions—will support this strategy. To be effective, each node (ground station or satellite) must be separated from another node by a large enough distance to prevent a single attack from killing more than one node.

TABLE 8-10. Space Survivability Hardening Design Summary. Though the space environment is harsh, survivability can be designed into spacecraft subsystems.

Threat Type

Requirement Driver

Mitigation Design Approach

Natural Space Radiation

Enhanced Raéatíon from Nuclear Bursts

Withstand total dose degradation. Minimize single-event upsets (SEU)

Radiation resistant materials, optics, detectors and electronics. Shielding at unit & part levels. Self-correcting features for SEU tolerance.

Collateral Nuclear Burst

Withstand prompt X-ray, neutron, EMP damage, minimize dose rate upsets. Tolerate induced noise due to debris.

Radiation resistant materials, optics, detectors & electronics. High Z shielding, current limiting/terminal protection. Event detection, circumvention, recovery. Sensor noise suppression. Multiple satellite coverages.


Sensor tolerance to background levels.

Processing algorithms. Multiple satellites for detection.

Ground Based Laser

Sensor tolerance to interference or damage.

2 color sensor detection, filtering and processing.

High Power Microwave and EMP

Sensor and communications tolerance to interference/damage.

Protection of detectors and circuits, processing for noise discrimination.

0 0

Post a comment