109 1010 10" 10« Huence (Particle • cm"2)

undergoes nuclear interactions within an electronic part Thus, lower-Z (and more abundant) ions deposit as much energy in a device as less abundant, higher-Z ions. When the galactic cosmic ray leaves electron-hole pairs in a depletion region of an electronic device, the electric field in that region sweeps up the pairs. Figure 8-14 shows this process schematically.

Fast Charged Particle

Energetic Proton

Fast Charged Particle

Energetic Proton

Depletion Region

Fig. 8-14. Schematic Showing How Galactic Cosmic Rays Deposit Energy In an Electronic Device.

Depletion Region

Fig. 8-14. Schematic Showing How Galactic Cosmic Rays Deposit Energy In an Electronic Device.

Single-event phenomena include three different effects in electronic components. The first is the so-called bitflip, or single-event upset (SEU), which neither damages the part nor interferes with its subsequent operation. The second is single-event latch-up (SEL). In this case, the part hangs up, draws excessive current, and will no longer operate until the power to the device is turned off and then back on. The excessive current drawn in the latched condition can destroy the device if the power supply cannot handle the current When latchup demands too much current for the power supply, it may drag down the bus voltage, or even damage supply. The third effect is single-event burnout (SEB). This causes the device to fail permanently.

To evaluate the frequency of single-event phenomena for a given part, we must know three things: the external environment; how the incident energy spectrum and particle intensity change as a particle passes through the spacecraft to the sensitive device; and how the electronic device responds to ionizing radiation. We find these phenomena difficult to evaluate because of the complex interactions between the radiation environment and the device's circuit elements. On-orbit failure rates can be predicted primarily for single-event upsets in memory devices, with well defined sensitive volumes, in which the galactic cosmic rays produce electron-hole pairs. A useful equation developed by Petersen [1995] expresses the upset rate R as follows:

where R is the number of upsets or errors per bit day, ai is the limiting cross section (sensitive area) of the device in cm2, and Lq ^s is the linear energy transfer (LET) at 25% of the limiting cross section in units of MeV/mg/cm2. If experimental cross section data is not available, but device modeling data is, then geometric data can be used in conjunction with the predicted critical charge, and now

where J? is the number of errors per bit day, a and b are device surface dimensions in pm, c is the device depth in pm, and Qc is the critical charge in pC. These two equations have been shown to predict upset rates in the geosynchronous orbit for solar piinimum conditions with reasonable accuracy. Scale factors for estimating error rates for other orbits and other calculational methods may be found in Petersen [1995].

Single-event upset rates in complex devices such as microprocessors, or single-event latchups or burnouts in any devices, cannot be reliably predicted. We must resort to predictions based on simulated accelerator test observations and flight performance of similar devices.

Galactic cosmic rays can also generate background noise in various satellite subsystems such as star sensors, infrared detectors, and components employing charge-coupled devices. In addition to increased noise signals, these rays create spurious events which can masquerade as real signals. The spurious signals can affect satellite subsystems depending on the genuine signals' frequency of occurrence, time duration, and repetition, as well as the sophistication of the sensor system. Galactic cosmic rays are a potential source of background noise which must be taken into account when designing a satellite system. It should also be noted that, while this section addresses effects of galactic cosmic rays, similar effects are caused by high energy protons and must be considered for orbits in the range of 1,000-10,000 km altitude.

8.2 Hardness and Survivability Requirements

Paul Nordin, The Boeing Company Malcolm K. Kong, TRW Systems & Information Technology Group

Survivability is the ability of a space system to perform its intended function after being exposed to a stressing natural environment or one created by an enemy or hostile agent. Hardness is an attribute defining the environmental stress level which a space system can survive. As an example, a satellite or spacecraft which can withstand an X-ray fluence of 1.0 cal/cm2 or absorption of 107 rads (Si) of total dose (a rad of absorbed dose is approximately 100 ergs/g) has a hardness of that amount (Fluence is the time integral of flux. Flux is the flow of energy per unit time and per unit cross-sectional area.)

In the aerospace industry we now consider both natural and hostile environments in the definition of hardness and survivability. Well-developed technologies, evolved over the last 35 years, make it possible to design satellites to withstand natural and modest levels of hostile environments. Although technologies for hardening against hostile military threats and for natural survival of satellites overlap, they are distinct and are usually treated separately except in the areas of survivability to total dose due to the Van Allen belts, single-event effects (SEE) caused by galactic cosmic rays and high energy protons, and space/bulk charging due to naturally occurring space plasmas. The latter phenomena must be treated synergistically in the design of satellites.

A military space system or commercial satellite must be survivable if we will need its services in times of high stress, such as a nuclear war. To do this, we must understand what may cause the system to malfunction and then design it to protect against failures. Survivability requirements include identifying the environments and their intensities and, in most cases, designing the space system so it will continue to perform its intended function for a specified time after exposure.

Commercial or scientific satellites usually do not need to be survivable to military threats, but planners must be aware that an unhardened satellite may prematurely stop operating after even very distant nuclear explosions. A slight hardening of satellites can make them much more survivable. (See Sec. 8.2.3.) The Starfish high-altitude nuclear test of July 9,1962, illustrates the vulnerability of unhardened satellites. That test, a 1.4 megaton device at 400 km altitude above Johnston Island in the Pacific Ocean, caused the failure of several satellites when electrons became trapped in the Earth's geomagnetic field. As a result of those failures, the U.S. Joint Chiefs of Staff established hardening guidelines for all military satellites, including operational and experimental ones. Ritter [1979] discusses these guidelines.

Studies were conducted by the Defense Threat Reduction Agency (formerly the Defense Special Weapons Agency) [Webb, et al., 1995] on the possible effects of a small Third World nuclear burst (for example, 50 kT at 120 km altitude over the East Asian Peninsula) on known commercial (unhardened) satellites. Satellites considered included Hubble Space Telescope, Iridium, ORBCOMM, Globalstar, NOAA, and Nimbus. These satellites showed lifetime reductions of 67% to as large as 99%.

It is important to consider survivability from the outset of mission design. For example, if the satellite can function within a range of orbit altitudes, the highest of these is both the hardest to attack and the most expensive to reach. We should consider the system's survival in each of its life-cycle phases, including concept definition, engineering design and development, and operations in orbit Note, however, that historically we have not hardened launch systems because of cost and weight as well as undefined need. The main military threats against space systems are nuclear weapons, including directed energy designs such as X-ray lasers; ground- and space-based laser weapons; high-velocity-pellet (fragmentation) weapons; high-power radio frequency (microwave) weapons; homing kinetic energy weapons; and beam weapons using neutral atomic particles. We may use several approaches to make a system survivable, with hardening of the satellite as a key element. Section 8.2.4 describes these possible approaches and discusses their approximate cost and relative effectiveness.

8.2.1 The Nuclear Weapons Environment and Its Effect on Space Systems

Nuclear weapons pose the most severe threat to spacecraft or space systems. The yield, or explosive power, and accuracy of delivery are such that if a nuclear weapon directly attacks a spacecraft, ground station, or any other node of a space system, the node will be destroyed. Nuclear weapon yields can range from a few tons to many megatons of TNT equivalent (one kiloton of TNT is defined to be 1012 calories). Future nuclear exchanges could use yields of a few hundred lalotons to a few megatons, depending on the purpose of the specific attack and the weapon's delivery accuracy. Accurate delivery of low yields will achieve the desired kill probability, whereas less accurate delivery requires higher yields.

Approximately 80% of the energy from a nuclear weapon detonated in space appears in the form of X-rays. Other important effects include small amounts of gamma rays and neutrons, as well as small fractions in residual radioactivity and kinetic energy of bomb debris. For additional technical detail on nuclear weapons effects, see Glasstone and Dolan [1977].

X-Radiation. The X-radiation occurs because just after detonation, nuclear bomb material is at 10-100 million K. As a first approximation, the hot bomb material will very quickly radiate the energy as though it were a black body, according to the Stefan-Boltzmann's law.

where E is the energy in W/m2, T is the absolute temperature in K, and a is the Stefan-Boltzmann's constant (5.67 x 10-8 W • nr2 • K-4). At higher black-body temperatures, more X-ray photons are emitted at higher energies. (This is Wien's law, which states that T= constant, where is the wavelength at maximum intensity and 7" is the absolute temperature of the blackbody.)

The X-ray fluence, Fx, at a distance R from a nuclear detonation of yield Y is given by

where fx is the fraction of the energy emitted as X-rays (■=» 0.8) and in the numerical form Fx is in cal/cm2, Y is in kilotons, and R is in km.

Neutron Radiation. One kiloton of equivalent nuclear energy arises from the fission of approximately 1.45 x 1023 nuclei. Each fission produces 2 or 3 neutrons. Approximately half of these neutrons escape during the few tens of nanoseconds of energy generation. Accordingly, the neutron fluence at a distance R cm from a nuclear detonation is given by

where Y is in kilotons, R is in km, and Fn is in n/cm2. This equation is only approximate. The actual neutron output will depend upon the design of the nuclear weapon.

Prompt Radiation. Gamma radiation emitted during the actual nuclear bum time is prompt radiation, whereas gamma rays emitted after the nuclear bum time are delayed radiation. Prompt gammas result from fission reactions, neutron capture, and inelastic neutron scattering events occurring during intense generation of nuclear energy. The total energy and energy distribution of the prompt gamma rays depend on the nuclear weapon's specific design. To calculate preliminary survivability at range R from a nuclear burst in space, we can express the dose, Dy, (energy deposited per unit mass) in silicon semiconductor material from prompt gamma radiation as

where ft is in km, Fis in megatons, and Dy is in rads (Si).

Delayed Radiation. Delayed gammas, neutrons, positrons, and electrons—or residual radiation—occur when radioactive fission products decay. For about the first second after a nuclear explosion, the decay rate from residual nuclear radiation is nearly constant Thereafter, the dose rate follows the approximate law:

where R is the dose rate (usually in rads/hr) at any elapsed time after the reference time, i0, when the dose rate was ft0- The fission products causing this dose rate contain more than 300 different isotopes of 36 elements from the periodic table, so the inverse 1.2 power of time is an approximation. It is accurate to within 25% for the first 6 months after the nuclear explosion [Glasstone and Dolan, 1977].

The explosion energy rapidly disperses residual radiation. In the absence of an atmosphere and geomagnetic fields, the radioactive fission products would expand geometrically and decrease in intensity by the inverse square of the distance from the burst However, the geomagnetic field causes the mostly ionized, radioactive weapon debris to spiral along geomagnetic field lines, in a manner similar to the charged-particle motion described in Sec. 8.1. Thus expansion of the radioactive debris will depend on the magnetic field at the nuclear event and on the magnetic field's configuration between the nuclear event and the satellite being considered. As a first approximation, we can assume a geometric expansion. A more conservative approach, however, would be to assume that the nuclear event and the satellite are on die same geomagnetic field lines. The largest possible amount of radioactive debris would then funnel from the nuclear event to the satellite.

An upper bound estimate of the delayed gamma flux due to radioactive debris, in gammas or photons per square centimeter per second, from a single nuclear burst, is given by

where Y is yield in megatons (one-third of total yield assumed to be fission), R is distance from the burst point in kilometers, and t is time after burst in seconds [Gold-flam, 1990]. This estimate applies to cases where the debris strikes and plates outer surfaces of a satellite, as well as cases where the debris is far away. Similarly, an estimate of delayed beta debris can be made by applying a one-third factor to the equation.

Both delayed gammas and betas will manifest themselves as noise spikes in electro-optical and visible sensor elements used on satellite systems (such as infrared surveillance sensors, optical/visible sensors, Earth sensors, and star trackers). The delayed gammas are a significant threat to satellites, since they can be reduced only by very thick shielding with high Z materials.

Electromagnetic Pulse (EMP). EMP is a secondary effect of nuclear weapon detonations. X-rays and gamma rays impinging upon the upper atmosphere create an electron flux which radiates in the RF region of the spectrum. EMP's spectral energy is mostly in the 1 MHz to 100 MHz range. As the RF energy arrives at a satellite, it will induce currents and voltages that may damage or kill the satellite if we do not design to protect it Nominal electric field strengths, which satellites would experience, can vary from 3 to 100 V/m, depending on satellite altitude and burst location, relative geometry, and other parameters.

System-Generated EMP, SGEMP, is a phenomenon caused when X-rays and gamma rays hit a satellite or other system element, thereby creating an internal flux of electrons whose electromagnetic interactions create large currents and voltages. These large internal currents and voltages can damage sensitive components inside the satellite. Section 8.2.3 discusses how we can mitigate these effects.

Geomagnetically Trapped Radiation. Following a nuclear burst at high altitude, electrons caused by the weapon join the naturally occurring Van Allen radiation belts (Sec. 8.1). The electron flux may increase by many orders of magnitude, thus increasing the absorbed dose in unshielded materials as the satellite repeatedly traverses the Van Allen belts. To protect solid-state (silicon) electronic circuits, we normally enclose them in aluminum, with wall thickness ranging from 0.0254 cm (0.01 inch) to a centimeter or more. Aluminum shielding with a thickness of 0.1 cm and a density of 2.71 g/cm3 corresponds to 0.27 g/cm2. Figure 8-9 of Sec. 8.1 gives the natural dose rate in silicon, in rads per year, for polar orbits as a function of orbital altitude and for several values of aluminum shielding. The dose scales linearly with time so the curves can be used for longer or shorter durations. A polar orbit satellite will accumulate less dose than an equatorial one because the trapped radiation is essentially nonexistent at and near the geomagnetic poles of the Earth; this dose difference can be as large as a factor of 5.

One Year Total Dose - 30 deg Inclinations

One Year Total Dose - 30 deg Inclinations

A. Natural total dose In one year, 30 deg Inclination circular orbits, for three values of aluminum shielding (0.254 cm, 0.508 cm, and 0.762 cm).
0 0

Post a comment