At least seven cycles have been studied for converting the thermal energy released by a nuclear reactor into electrical power for spacecraft applications.(1) Three of these, the Rankine, Brayton and Stirling cycles, involve dynamic conversion, while four others, Alkali-Metal Thermoelectric Converter (AMTEC), thermoelectric, thermionic, and magnetohydrodynamic, are static systems.
The operation of dynamic conversion systems may produce vibrations which could compromise spacecraft payload operations, a problem avoided by static systems (Figure II-21). But the static thermoelectric and thermionic conversion systems are inherently less efficient than the dynamic cycles (Figure II-22)(2). While magnetohydrodynamic is highly efficient, in practice it is typically operated in an open mode, which may produce exhaust effluents which may interfere with payload operations. Choice of the appropriate conversion cycle requires a detailed evaluation of spacecraft and payload requirements and constraints.
Each of these cycles has attractive operating features at different temperature ranges. Although thermal efficiency is one commonly used figure of merit, high theoretical efficiencies are seldom realized in practice, given other operational constraints on system design.
Direct One Loop
Indirect Two Loop
Indirect Three Loop
1 - Rankine Cycle
The Rankine Cycle is based on a working fluid that changes phase during the operation of the conversion cycle. The most common terrestrial application of this cycle is in coal-fired and nuclear electrical power plants, in which the working fluid is water which is heated to form steam which drives turbine generators. The boiling and condensation phases allow addition and extraction of heat at constant temperature, which allows this cycle to approach ideal Carnot efficiency. The circulation of the working fluids is a fairly efficient process which does relatively little to detract from the efficiency of the cycle's operation. There are two variants of this cycle:
Liquid Metal systems use metals with low melting and boiling temperatures, such as mercury, cesium, potassium, sodium or lithium (in increasing order of operating temperatures) as working fluids.
Organic Rankine systems use organic liquids such as terphenyl, pyridine (CP-32), monoisopropyl biphenyl (MIPB) or a eutectic of biphenyl and biphenyl ether (Dowtherm A). The potential for pyrolytic degradation of these fluids restricts their use to operating temperatures considerably below those of liquid metal systems.
There are two alternative implementations of this cycle:
Direct Rankine systems consists of a liquid (typically potassium in liquid metal reactors) that is vaporized by the reactor or other power source, passes through turbomachinery to produce electric power, and is then condensed back to the liquid phase for recirculation through the system.
Indirect Rankine systems use a liquid (typically lithium in liquid metal reactors) that is heated by the reactor or other power source, which then passes through a heat exchanger to vaporize a working fluid (potassium in liquid metal reactors) in a secondary loop, which in turn is passed through turbomachinery and a condenser. This inclusion of a heat exchanger in implementation of the Rankine cycle adds additional weight and complexity compared to the Direct cycle, but avoids the difficulty of assuring in-reactor liquid-vapor separation in the microgravity environment of space.
The primary advantage of the Rankine Cycle is high efficiency, and isothermal heat rejection, which minimizes the area of a radiator for a given operating temperature. This system can provide useful levels of power output at relatively low operating temperatures, with radiators operating at relatively high temperatures, compared to the Brayton Cycle.
The primary drawback of the Rankine Cycle is the corrosive and erosive effects of the liquid metal working fluids used. In particular, the use of lithium poses a risk of fire or explosion, should leaks in the coolant loop expose the lithium to ambient water contaminants.
In addition, the simplest Rankine cycle, which would use a single loop circulating the working fluid from the reactor through the power turbines and the radiators, could expose the entire conversion system to the radioactive potassium coolant. Rankine cycle systems use two coolant loops connected by a heat exchanger, with the secondary loop circulating non-activated fluid through the power turbines and radiators. A third loop may also be added to drive hydraulic systems (Figure II-23). These additional loops add to the mass and complexity of Rankine systems.
The fact that the Rankine cycle is based on the use of working fluids that are present in both liquid and vapor phases adds a further element of complexity. In-core boiling raises concerns about critical heat flux and reactivity effects, as well as about erosion and corrosion. Particular attention is required to issues such as control of liquid-vapor interfaces and achieving adequate liquid vapor separation, as well as management of fluid transport and fluid inventory location. While these matters are resolved with relative ease in terrestrial systems, their resolution in the microgravity environment of space can provide a challenge to system designers.
2 - Brayton Cycle
The Brayton Cycle uses a single-phase gaseous working fluid which absorbs energy at a constant pressure through a rise in temperature. The Brayton Cycle may be operated in either on open or closed mode.
Open Brayton Cycle is used for terrestrial applications, such as gas-turbine propulsion systems used on many naval vessels, or in fixed-site auxiliary electrical power generation systems. In this cycle, the gas is exhausted into the environment after passing through the power generating turbo-machinery.
Closed Brayton Cycle passes the gaseous working fluid through power generating turbo-machinery, after which it is circulated through a radiator for cooling and reuse. The liquid fluids used in the Rankine cycle have heat-transfer coefficients approximately 50 times higher than the gaseous fluids used in the Brayton cycle. As a result, the closed Brayton cycle has an inherently lower thermal efficiency than does the Rankine Cycle. The efficiency of this cycle is further compromised by the work required in compressing the large volumes of high pressure gaseous working fluid. As with the Rankine cycle, more than one fluid circulation loop may be used in Brayton cycle systems.
The Brayton Cycle has several attractive features:
+ Brayton systems have a larger experience base and greater technological maturity than other conversion systems;
+ The simplicity of the single loop circulation of the single phase gaseous working fluid;
+ Open-cycle cooling can be more readily implemented;
+ The non-corrosive properties of inert gases which may be used as working fluids.
+The cycle can accommodate higher source temperatures than can the Rankine cycle.
+ Brayton systems can readily respond to rapid changes in power load requirements by adjusting operating pressure, while Rankine systems require longer response times due to coolant thermal inertia.
However, the Brayton Cycle has several drawbacks, resulting from its low inherent thermal efficiency:
+ Gas-cooled reactors require larger heat transfer surface areas than do liquid-metal cooled Rankine reactors, and as thus typically require larger and more massive reactors.
+ The cycle requires very high efficiencies in power and compressor turbine components;
+ Inert gas working fluids may become contaminated with corrosive impurities, compromising system operations;
+ Very high operating temperatures and pressures may be required to achieve satisfactory power outputs;
+ High rotational speeds of primary generation turbines may require the use of either specially designed high-speed electrical generators which may add to cost and technical risk, or complex reduction gearing assemblies with add mass and compromise efficiency.
+ For a given heat source temperature, the Brayton cycle requires a larger radiator area compared to Rankine systems.
+ Thus Brayton systems are typically more massive than other cycles.
3 - Stirling Cycle
The Stirling cycle is a closed-cycle reciprocating engine that uses a high-pressure single-phase gaseous working fluid, such as hydrogen or helium. The primary advantage of this cycle is its high thermodynamic efficiency, approaching theoretical Carnot efficiencies, at low operating temperatures. However, despite development efforts over nearly two decades, this technology has yet to demonstrate a firm technology base for long-life operations. In addition, most of the research in this field has concentrated on kilowatt-class applications, which may not be readily scaleable to megawatt-class operations. The alternative, use of large numbers of small engines, would make this cycle too heavy for use for high-power applications.
4 - Alkali-Metal Thermoelectric Converter (AMTEC)
The Alkali-Metal Thermoelectric Converter (AMTEC) cycle, which is thermodynamically similar to the Rankine cycle, uses high-pressure sodium vapor which is supplied to one side of a beta alumina solid electrolyte. The transport of sodium ions across this electrolytic membrane produces a voltage differential, with low pressure sodium vapor emerging from the other side of the membrane. This voltage drives electrons through a useful load prior to their completion of the circuit with the sodium vapor.
5 - Thermoelectric Conversion
Thermoelectric systems rely on the direct conversion of thermal energy to electricity by means of the Seebeck Effect, which was discovered in 1822. The Seebeck effect refers to the obervation that a temperature differential between two adjacent, dissimilar materials generates a voltage between them. Interest in the practical applications of this phenomena awaited identification of materials which exhibited interesting levels of conversion efficiency. The development of semiconductors in the 1950s led to the fabrication of a range of intermetalic compounds and alloys which produced conversion efficiencies of several percent. These low efficiencies, coupled with the thermal properties of these materials, have limited the applications of thermoelectric conversion systems to relatively low power levels. Thus thermoelectric converters are frequently used in Radio-isotope Thermoelectric Generators (RTGs), which provide up to a kilowatt of power for deep space probes.
Candidate thermoelectric converter materials include PbTe, Pb-Sn-Te, and various Ge-Si alloys. These materials have theoretical conversion efficiencies of 10% to 15%, although materials characteristics may limit operating temperatures to ranges that will result in lower efficiencies. Although the mass of the converter system itself is fairly small, thermoelectric power systems also require relatively massive reactors and radiation shielding components. Because of the inherently low conversion efficiency, and the material temperature limits which further reduce efficiency, larger radiator areas are required. Thus in practice, design considerations typically limit thermoelectric systems to net efficiencies of about 3%.
6 - Thermionic Conversion
Thermionic conversion systems may be conceived as an "electron boiler" in which electrons are thermally boiled off a heated emitter cathode, and collected on an anode surface, delivering electrical power to an external load.
At least three configurations have been suggested for thermionic nuclear power systems:
+ A cylindrical reactor core with thermionic converters on its outer surface, with each converter having its own radiator. This configuration is limited to low power levels by its limited emitter area.
+ Out-of-core configurations can achieve higher power densities and larger emitter areas by circulating a liquid metal coolant through the reactor core, which transfers the heat to an external panel of converters. While in all configurations the reactor is operating at a higher temperature than the emitter, this design requires the coolant loop to operate at the temperature of the converter emitter cathode, which poses significant design challenges.
+ In-core thermionic systems are composed of converters which are directly attached to individual fuel elements within the reactor core. Anode cooling is by a loop which circulates a coolant (typically using a liquid metal) to an external radiator.
Conversion efficiencies increase with emitter temperature, with theoretical efficiencies ranging from 5% at 900o K to over 18% at 1,750o K. In practice, efficiencies may range from less than 4% at 900o K to over 9% at 1,750o K. Uranium Oxide (UO2) fast reactors are attractive for maintaining the high temperatures and high power densities required for in-core thermionic systems.
Attractive features of thermionic systems include:
+ Static operations for reduced vibration and improved reliability;
+ High power density;
+ Potentially high thermal conversion efficiency compared to thermoelectric systems;
+ Very high heat-sink temperatures reduce radiator mass penalties;
+ Liquid-metal (NaK) cooled thermionic reactors operate at lower temperatures than other metal-cooled concepts, and thus face reduced corrosion concerns.
However, the high core power density, high operating temperatures and intense radiation environments of Thermionic reactors pose significant material selection and configuration design challenges to provide both high efficiency and high power with operating reliability over extended periods. Fuel element swelling, leading to structural deformation or radiation leakage may be a particular problem. Thermionic reactors using lithium or sodium-potassium coolants also run risks of accidental fires. One analysis noted:(3)
"Fuel fabrication for thermionic reactors is much more difficult than for other reactor types since the fuel element includes an emitter, collector, insulators, spacers, sheath, cladding and fission product purge lines. All these components must be held to tight tolerances. For advanced thermionic concepts, this problem will be more severe because thousands of very small diodes will be required."
A primary concern with Thermionic systems is their ability to maintain satisfactory operations over extended periods. While experience to date has demonstrated operating lifetimes of a few years, a two-to-three fold improvement may be required for some applications. The Thermionic Fuel Element (TFE) test program was initiated to attempt to resolve some of these issues.
6 - Magnetohydrodynamic (MHD) Conversion
Magnetohydrodynamic (MHD) conversion systems produce electrical power according to the same principles used in conventional solid-core generators, except that the rotating magnet of the conventional generator is replaced by an electrically conductive ionized plasma, which passes through a channel surrounded by a magnetic field which is at right angles to the motion of the plasma (Figure II-24). The linear motion of the conducting gas
is substituted for the rotational motion of the metallic conductor used in a conventional generator. The passage of the plasma through the generator's external magnetic field produces an electrical current according to the principles first observed by Faraday in 1831. The power generated is proportionate to the channel volume, plasma velocity, and the field strength of the surrounding magnets. Losses are proportionate to channel area.
Advantages of the MHD conversion system include:
+ High enthalpy extraction, which at high temperatures may surpass 50% conversion of thermal energy to electric power in open cycle systems(4) (although in practice efficiencies are typically in the range of 10% to 15%).(5)
+ Few or no moving parts, which can improve reliability compared to dynamic conversion systems, as well as reduce vibration concerns.
+ Simplicity of operations facilitates long-term storage in space with only infrequent testing or operations.
+ Favorable scaling to larger sizes.
+ Very short, essentially instant, turn on, which is important for burst-mode applications.
+ High power densities, in the range of 5 - 7 kW/kg, though net power densities may range below 2 kW/kg.(6)
The primary disadvantage of MHD systems is the large volume of high temperature effluents created during operations. The hot gas passing through the generator, which may be hydrogen in the case of nuclear systems, or water vapor or other molecules in the case of chemical rocket systems, must be seeded with conducting materials such as lithium, potassium or cesium. Although cesium is more expensive than potassium, it offers higher power densities. These seed materials, as well as other effluents from chemical power systems, can interfere with payload operations if directly exhausted into the space environment by an open cycle MHD generator. Although closed cycle systems avoid this problem, they require large and massive effluent treatment and absorption systems, which can significantly compromise the attractiveness of this system.
Removal of cesium seed material from the MHD effluent stream of open cycle systems will require a significant development effort:(7)
"The current state of the art of cesium removal from the exhaust stream is based on water cyclone scrubbers to condense and remove the alkali metal from the gas. This approach has been used extensively to remove the alkali metal (potassium or cesium) from many laboratory MHD installations. The requirement... is to perform the same task in space using a fixed amount of hydrogen available from the weapon cooling system. The fluid mechanics and energy balance are well defined for these requirements. However, the difficulty of the task is to accomplish it with a minimum mass and volume and a fixed amount of hydrogen."
One analysis of multimegawatt space power systems observed:(8)
"In practice, introducing MHD technology poses several practical problems in addition to its extremely high operating temperatures and the need to obtain adequate electrical conductivity. One category of problems relates to achieving satisfactory behavior of the fluid flow in the MHD channel during the conversion process. Another problem is the management of system effluents emerging from the channel.
"Mitigation of the first problem requires attaining a highly ionized, high-velocity gas stream having adequate uniformity. The gas flowing through the MHD channel consists of a mixture of hot combustion products of an exothermic reaction -- which provide fuel -- seeded with an alkali metal (eg potassium) to improve electrical conductivity when ionized. Small nonuniformities of gas density and/or ionization concentration (conductivity) can result in major flow instabilities, and the excess heating in these regions causes acoustic disturbances and flow disruptions.
"The effluent problem requires finding channel geometries that maximize uniformity of flow and minimize excess heating -- and the resultant acoustic disturbances -- in the conversion and exhaust regions of the channel. The high-MHD generator-exhaust temperatures (about 2500o K) pose difficult materials problems and, if the escaping ionized gas is discharged into space, the glow emitted by the recombination of its ions and electrons would be detectable to an enemy.
This analysis identified:(9)
"... problem areas such as uniformity of the ionized gas in the MHD conversion channel, channel erosion, and dealing with substantial quantities of metallically seeded ionized effluents. MHD space power systems could degrade spacecraft stability or perturb orbits.
Based on these challenges, the panel concluded:(10)
"... until MHD systems that might be developed for SDI are projected to be capable of modifying or trapping such effluents... further MHD development for SDI -- beyond the conceptual studies and scaling validations presently contemplated -- is not warranted."
7 - Summary
Despite four decades of analysis and development, choices among conversion cycles remain complex and contentious. For many of these cycles, theoretical advantages are offset by severe engineering challenges in implementation, while for other cycles, comparative ease of realization may adequately compensate for low theoretical efficiencies. No ready choice is apparent between closely competitive technologies, and the debates between the proponents of Rankine and Brayton cycles, and the advocates of Thermoelectric and Thermionic systems, are at least as contentious today as they were in the 1960s.
While in theory one might hope that there would be clear discriminants among conversion cycles, in practice this does not appear to be the case. Given the disparate range of figures of merit on which systems may be evaluated, proponents of each system are able to point to at least some metrics which support the superiority of their favored approach.
The naive expectation that each conversion cycle would have a clear-cut domain of superiority founders on the reality of the complexity of choosing among widely dissimilar approaches to common problems.
These perplexities notwithstanding, some general observations are possible:
+ None of the proposed conversion cycles has demonstrated reliable long-life operations at multimegawatt power levels -- all require significant development efforts to reach these goals.
+ Some conversion cycles, notably Stirling (for lower power applications), and AMTEC and MHD (for multi-megawatt applications), require major technology improvements, given relatively limited terrestrial experience base.
+ In the near-term, the Brayton cycle is the preferred approach for burst-mode open-cycle systems, and appears competitive with the Rankine cycle for closed-cycle steady-state systems.
+ In the longer run, the Brayton cycle will remain the system of choice, absent major improvements in MHD conversion, for open cycle systems. But technology improvements may render the Rankine cycle preferable to the Brayton cycle for closed systems.
E - RADIATOR CONCEPTS
The choice between open and closed cycle power systems remained one of the central issues of the Multimegawatt program. For open cycle systems, the dominant scaling factor is the duration of operations, with extended operations requiring increasing amounts of expendable coolants (Figure II-25). For close cycle systems, the primary consideration is the increasing radiator mass needed to accomodate higher power levels (Figure II-26).
The selection of either open or closed cycles has direct implications for the choice between chemical or nuclear power sources. Although nuclear systems could be operated in either an open or closed mode, chemical power sources were typically more compatible with open-mode operations, except for long-duration operations (Figure II-27).
The effects of effluents from open-cycle operations on spacecraft and payload operations is perhaps the most significant issue facing such systems.(11)
"The management of effluent from open cycle power systems is a primary issue of concern. In order to prevent effluent from the power system from interfering with the operation of sensors and directed energy weapons, it is important to employ high velocity directional nozzles which produce highly directed jets of the effluent into areas of the space environment out of the field of view required by the payload... flow from such nozzles must have a high Mach number, M = 5 or greater, in order to achieve a relatively clear view for the spacecraft. Even under these conditions, a significant amount of effluent back streaming into critical view areas may occur."
"The thrust produced by a high Mach number nozzle for effluent discharge can be significant in the case of an SDI weapons platform. Estimates of these thrusts are in the range of 104 - 105 N... in order to avoid thrusting the vehicle, most designs use configurations such as an opposed group of nozzles in a dipole or quadrupole configuration... While eliminating net thrust, these designs increase the amount of back streaming into the field of view for the weapon systems sensors, and they require precision balancing of the opposed nozzles in order to avoid slewing or jittering the spacecraft."
"An early concern about open cycle spacecraft was interference with a Neutral Particle Beam. Preliminary indications show that stripping of the beam is not likely to occur if the effluent is composed only of neutral hydrogen, and if highly directed effluent nozzles are employed at some distance away from the NPB output. Nevertheless, a concern remains regarding ionized effluent which could be produced by radiation, Van Allen interaction, and collisions between the effluent gas molecules and residual atmosphere of the earth (known as "ram")... The ionized effluent could become trapped around the spacecraft by fringe magenetic fields from the beam expander telescope and beam director magnets."
"... X-ray flashed from nuclear bombs can result in much more intense ionization of the effluent cloud. Ionization from a nuclear weapon could tend to be of short duration and, again, is more likely to be a problem if the plasma which is produced around the spacecraft is trapped rather than dissipated."
"For a space platform at relatively low altitudes, the orbital speed through the residual atmosphere creates an intense ram effect, which can heat (an possibly ionize) an effluent gas cloud... These ram effects are important at an altitude of 400 km and lower. Ram would not be significant at altitudes >1000 km unless a local disturbance such as atmospheric heave (upwelling caused by a nuclear explosion) or effluent emitted by another spacecraft produces a temporary gas cloud at high altitude."
From the outset, it was recognized that meeting SDI MMW power requirements would entail significant advancement in radiator systems:(12)
"Compared to existing systems, significantly higher power production and rejection rates demand the development of radiators and power conversion systems with very low specific weights."
The challenge of radiator design is not limited solely to accomodating power source thermal management, since:(13)
"... there is still the problem of removing the waste heat from the currently popular weapon concepts in SDI systems. A closed cycle radiator for the weapon would typically be much larger than that of the power system, because the weapon waste heat is rejected at a much lower temperature."
Although a variety of reactor core and conversion cycles have been evaluated under the Multi-Megawatt program, radiators are perhaps the most critical technology for closed-cycle systems. At high (megawatt) power levels, the mass of the radiator becomes the dominant factor in the total mass of the power system. This is a key consideration in favor of the utilization of open cycle systems where possible, and the development of more advanced radiator concepts for use in applications which require closed cycle power sources.
An additional concern posed by advanced radiator concepts is their relative technical immaturity. In addition, choices among radiator concepts are complicated by a range of potential figures of merit. One analysis noted:(14)
"The present competition between radiator systems concentrates on specific weight. This emphasis may be misplaced... Quite often, factors dealing with deployability, survivability, geometry, maneuverability, etc, become determining considerations."
1 - Fin/Tube Radiators
The primary advantage of these conventional radiator designs is their simplicity and the extensive existing technology base. However, for multimegawatt applications, they would require very massive radiator panels.
The mass of radiator panels is increased by the need to protect the coolant loop from punctures by micrometeroids and space debris (Figure II-29). Additional protection may be required from hostile actions as well.
A variant of this class is the Heat Pipe radiator (Figure II-30). This concept operates on the basis of an evaporation-condensation cycle, in which the gaseous circulating coolant is returned to the liquid phase as it passes through a heat exchanger connected to the radiator. The high effective thermal conductivity associated with this process results in a relatively low mass radiator system. The thickness and mass of the radiator surface is determined by puncture hazard considerations, rather than thermal conduction requirements. Since each heat pipe element of the radiator is an independent unit which is not connected with the primary thermal management coolant loop, damage to a single element by micro-meteoroid or space debris colision, or hostile attack, would not result in overall system failure.
Although heat pipes have been used for low-power thermal management on spacecraft, there is no flight experience base with systems for prime power thermal management. One outstanding technical issue is the problem of establishing efficient thermal linkage between the primary coolant loop and the heat pipe evaporators. Development of appropriate thermal coatings represents an additional challenge. In addition the large panels required by this system may pose difficulties in erecting panels in space.(15)
2- Membrane Radiators
Membrane radiators consist of a thin flexible membrane inflated by low gas pressure, with coolant circulating on the inner surface of the rotating membrane. These radiators may use either spherical (Figure II-31) or disk (figure II-32) membranes. In this concept:(16)
"The working fluid, a liquid or preferably a condensable vapor, is introduced inside the sphere and impinges on the top and bottom surfaces. The fluid wets the inner surface of the sphere and is driven in the form of a liquid film by centrifugal force to the equatorial periphery of the sphere; liquid metal pumps located there return the liquid out of the sphere through rotated shaft seals to its source. As the liquid flows along the inner surface of the envelope it loses heat by thermal radiation from the outer surface of the balloon."
The use of thim membrane materials in this radiator configuration offers significant reductions in system mass compared to conventional panel technology.
However, they are vulnerable to single-point failures, if the membrane is punctured, either
as a result of meteoroid or space debris collisions, or as a result of hostile actions. At low operating temperatures, surface tension of the working fluid could prevent leakage from small punctures. However, at higher temperatures the vapor pressure of the coolant increases, resulting in evaporation losses through punctures.
A further design challenge for this concept lies in the implementation of the various rotating machinery elements, which could pose reliability concerns for long-life systems. The rotating vapor seal between the radiator and the body of the spacecraft may prove a particularly problematic element. The feasibility of this concept ultimately depends on development of strong, light and rugged materials for the membrane envelope.
3- Belt/Filament Radiators
Belt and filament radiators are based on mechanical heat transport away from a heat exchanger by moving a thin ( 1 mm ) flexible material, which radiates waste heat prior to return to the exchanger. The thermal load may be imparted to the moving radiator either by direct contact (Figure II-33) or by convection (Figure II-34). An alternative approach substitutes a band of filaments for the rotating belt (Figure II-35)
A significant advantage of this class of radiators is the reduced vulnerability to damage from micro-meteoroids, space debris, and hostile attack, since the exposed radiator belt or filaments are mechanically robuts, while the more fragile heat exchanger is protected from exposure to these hazards. The belt may be folded prior to use, reducing storage concerns.
Although belt and filament radiators are potentially less massive than conventional designs, they employ complex operational concepts which could pose significant testing and reliability problems. In addition, the large flexible belts would complicate manevering and slewing of military space platforms.
4 - Dust/Droplet Radiators
Dust and droplet radiators operated by recirculating free streams of sub-millimeter particles to radiate waste heat directly into space.(17) Interest in these concepts originated in the late 1970s in response to concerns about the vulnerability of conventional radiator designs to damage by micrometeroids.(18) The initial concept used solid dust particles, but problems in management of the particle inventory, as well as the low efficiency (except at high temperatures) of heating the radiator particles (via radiation rather than conduction) limited interest in this approach Figure II-36). Thus subsequent investiagations focuse on more readily manipulated liquid droplet configuraitons.(19)
The primary advantage of these systems is a very large surface to mass ratio of the radiating surface compared with more conventional radiators, which could potentially result in an order-of-magnitude reduction in thermal management system mass. Although the total mass of the radiator system may be 4 - 8 times that of the droplet array itself, this concept can still produce significant mass savings relative to more conventional systems. Additional attractive features of this class of radiators include:
+ Reduced vulnerability to micro-meteoroid, space debris, and hostile attack than conventional systems;
+ Low transport volume;
+ Ease of deployment and stowage.
Major developmental challenges facing this class of radiators involve the complex hydraulics of the systems, including:
- Improvement in generation and precision direction of particle streams is required to reduce losses from misdireciton and collision, which could produce contamination interfering with payload operations;
- Charging of droplets due to spacecraft charging may result in significant stream deflection and droplet losses.(20)
- Improved efficiency in particle stream collectors, to reduce losses and contamination, may pose serious challenges in the microgravity environment of space;
- One of the most serious problems results from surface evaporation losses from droplets, which may constrain maximum operating temperatures to below 500 K. Evaporation could require unacceptably massive coolant reserves for replentishment, and could produce contamination interfering with payload operations;
- Radiative performance of radiating particle streams may be relatively low. While silicone and lithium have satisfactory thermal properties, their use in constrained by evaporation losses. the emissivity of liquid metals (such as tin) that might be used in higher temperature systems is typically of the order of 0.1, resulting in an effective radiator emissivity of 0.2, in constrast to the 0.85 emissivity typical of aluminum used in fin radiators.(21)
- Liquid droplet radiators using liquid metals such as tin could face very complicated shut-down, dormancy and restart problems with freezing and thawing the metal coolant.
Concerns about droplet losses were the basis for proposals to used electrostatically charged droplets or particles, that would be confined by the field lines established by charged generator and collector surfaces (Figures II-39, 40). But subsequent analysis disclosed that the charging of the droplets would be negated by ambient space plasmas, raising doubts about the practicality of this concept.(22)
A further variant on this concept is the Curie Point radiator (Figures II-41, 42).(23) In this system, solid ferromagnetic metal particles are heated above their Curie Point and ejected in a stream into space. As these paramagentic particles radiate heat, their temperature eventually drops below their Curie point, whereupon a magnetic field guides the now-magnetic particles into a collector.
While the Curie Point radiator may potentially offer a least-mass solution to thermal management, it employs unproven concepts, with complex and potentially massive support systems. In particular, the large magnets required by this concept may result in systems with specific masses comparable to traditional radiator concepts.
5- Enhanced Survivability Concepts
While the previously discussed radiator concepts are primarily intended to optimize performance figures of merit such as mass or efficiency, other radiator concepts are more focused on improving survivability.
A number of survivable radiator concepts have also been discussed.(24)
"... some closed cycle burst power systems use deployable radiators, which can be retracted during the dormancy period. Although this improves the survivability during dormancy, the lightweight deployable radiator may still be very vulnerable to laser and nuclear x-ray attack when it is deployed during operation. Balloon type radiators and collapsable heat pipes are examples of these concepts."
Another example of a survivable radiator is depicted in Figure II-43. In this concept, waste heat is radiated directionally, away from the Earth, to minimize the thermal signature of a spacecraft to lower altitude or terrestrial infrared sensors.
6- Performance Comparisons
The performance of a radiator system is a function of the emissivity of the radiator surface material, and the radiator's surface temperature. Radiator surface materials are generally chosen to optimize emissivity, and only marginal improvements are possible beyond current practice.
But the heat radiated from a surface is proportional to the fourth power of the surface temperature, so relatively small increases in surface temperature can produce very large improvements in radiator performance.
One analysis concluded:(25)
"Some low-mass concepts for heat rejection are worthy of consideration for heat rejection at low temperature (i.e. below 1000o K). These include the liquid droplet and liquid sheet radiator concepts, and various moving belt (liquid, solid, and hybrid) radiators. All of these concepts are in the advanced conceptual stages of development, and none of them have been adequately tested. Questions regarding maneuverability issues, particularly for the belt radiators, and contamination caused by escaping fluids have not been addressed. It is anticipated that such systems, if successful, would not be available before the year 2000. Nonetheless, the heat rejection issue is sufficiently critical that such advanced concepts merit consideration for future SDI systems."
1. William, R.A., et al, Nuclear, Thermal and Electric Rocket Propulsion -- Fundamentals, Systems and Applications, AGARD/NATO First and Second Lecture Series, 12-21 November 1962 and 28 September - 2 October 1964, (New York, Gordon and Breach, 1967), is the primary source for the following discussion.
2. Mullin, J.P., et al, "NASA's Space Energy Technology Program," Intersociety Energy Conversion Engineering Conference, 1984, paper 84-9388, page 548, and
Shmidt, J.E., et al, "A Liquid Metal Cooled Reactor - Alkali Metal Thermoelectric Space Power Systems Concept for Mult Megawatt Applications," Transactions of the Third Symposium on Space Nuclear Power Systems, Albuquerque, NM, 13-16 January 1988, page MM-3.1.
3. Department of Energy, Office of Defense Energy Projects and Special Applications, Strategic Defense Initiative Multimegawatt Space Nuclear Power Program - Summary, April 1986, page 26.
4. "SDI Turbo Saved From Davy Jones," Military Space, 27 May 1985, page 1.
5. Christensen, L.S., et al, "Updated Results of Large Scale MHD Generator Experiments," 17th Intersociety Energy Conversion Engineering Conference, 1982, paper 82-9206, pages 1217-1222.
6. Schmidt, H.J., et al, "Pulsed MHD Generator Development in the Soviet Union," 19th Intersociety Energy Conversion Engineering Conference, 1984, paper 84-9425, pages 1476-1484.
7. Martin Marietta, op cit, page 514.
8. National Research Council, Energy Engineering Board, Committee on Advanced Space Based High Power Technologies, Advanced Power Sources for Space Missions, (Washington, National Academy Press, 1989), pages 33-34.
9. ibid, page 34.
10. ibid, page 34.
11. Britt, Edward, et al, "Discussion of Open Versus Closed Cycle Space Power Burst Energy Systems," Transactions of the Sixth Symposium on Space Nuclear Power Systems, Albuquerque, NM, 8-12 January 1989, pages 357-362.
12. Department of Energy, Office of Defense Energy Projects and Special Applications, Strategic Defense Initiative Multimegawatt Space Nuclear Power Program - Summary, April 1986, page 10.
13. Britt, op cit, page 360.
14. Begg, Lester, and Wetch, Joseph, "Comparison of High Temperature Heat Rejection Concepts to System Related Requirements," 22nd Intersociety Energy Conversion Engineering Conference, Philadelphia, PN, 10-14 August 1987, volume 1, pages 227-234.
15. Mattick, A.T., and Hertzberg, A., "Advanced Radiator Systems for Space Power," 38th Congress of the International Astronautical Federation, Brighton, UK, 10-17 October 1987, paper IAF-87-230.
16. Koening, Daniel, "Rotating Film Radiators for Space Applications," Society of Automotive Engineers, 1985, paper SAE/P-85/164, 85-9401.
17. Mattick, A.T., "Experimental Test of Liquid Droplet Radiator Performance," Transactions of the Third Symposium on Space Nuclear Power Systems, Albuquerque, NM, 13-16 January 1986, pages TM-1.1 - 2.
18. Mattick, A.T., and Hertzberg, A., "Advanced Radiator Systems for Space Power," 38th Congress of the International Astronautical Federation, Brighton, UK, 10-17 October 1987, paper IAF-87-230.
19. Mattick, A.T., and Hetzberg, A., "Liquid Droplet Radiators for Heat Rejection in Spcae," Journal of Energy, November 1981, vol. 5, num. 6, page 387-393.
20. Mattick, A.T., and Hetzberg, A., "The Liquid Droplet Radiator: An Ultralightweight Heat Rejection System for Efficient Energy Conversion in Spcae," Acta Astronautica, 1982, vol. 9, num. 3, page 165-172.
21. Wetch, Joseph, et al, "Static and Dynamic High Power, Space Nuclear Electric Generating Systems," 1985, SAE Paper P-85/164, 859233.
22. Mattick, A.T., and Hertzberg, A., "Advanced Radiator Systems for Space Power," 38th Congress of the International Astronautical Federation, Brighton, UK, 10-17 October 1987, paper IAF-87-230.
23. Carelli, Mario, et al, "The Currie Point Radiator," Chapter 45 in El-Genk, M.S., Space Nuclear Power Systems 1988, (Orbit Book Company, Malabar, FL, 1989), pages 367-375.
24. Britt, Edward, et al, "Discussion of Open Versus Closed Cycle Space Power Burst Energy Systems," Transactions of the Sixth Symposium on Space Nuclear Power Systems, Albuquerque, NM, 8-12 January 1989, page 360.
25. National Research Council, Energy Engineering Board, Committee on Advanced Space Based High Power Technologies, Advanced Power Sources for Space Missions, (Washington, National Academy Press, 1989), page 29.