تجزیه و تحلیل عملکرد مبدلهای حرارتی آبشاری برای سیستم های قدرت رادیوایزوتوپ پیشرفته

Advanced radioisotope power systems (ARPSs) for future planetary missions require higher conversion efficiency than the state-of-the-art (SOA) SiGe thermoelectric converter in order to decrease system mass and reduce mission cost. The performance of three cascaded thermoelectric converters (CTCs) for potential use in ARPSs is investigated at heat rejection temperatures of 375, 475 and 575 K and input thermal powers of 1, 2 and 3 Wth. These CTCs have top SiGe unicouples that are thermally, but not electrically, coupled to bottom unicouples having one of the following compositions: (a) TAGS-85 (p-leg) and 2N–PbTe (n-leg); (b) CeFe3.5Co0.5Sb12 (p-leg) and CoSb3 (n-leg); and (c) segmented p-leg of CeFe3.5Co0.5Sb12 and Zn4Sb3 and n-leg of CoSb3. The top and bottom unicouples in the CTCs are of the same length (10 mm), but the optimized cross-sectional areas of the n- and p-legs for maximum efficiency are different. The nominal hot junction temperature of the top SiGe unicouples at their peak efficiencies is 1273 K and that of the cold junction is 780 K when the bottom unicouple is of composition (a) and 980 K for compositions (b) and (c). The hot junction temperatures of the bottom unicouples are taken 20 K lower than the cold junctions of the top unicouples, but the input thermal powers to the former are the same as those rejected by the latter. Assuming zero side heat losses and a contact resistance of 150 μΩ cm2 per leg in the top and bottom unicouples, the calculated peak efficiencies of the CTCs vary from 9.43% to 14.35%. These efficiencies are 40–113% higher, respectively, than that of SOA SiGe (∼6.5%) when operating at the cold junction temperature of 566 K and the same hot junction temperature (1273 K) and contact resistance per leg. Decreasing this resistance to a realistic value of 50 μΩ cm2 per leg increases the peak efficiencies of the CTCs by 0.5–0.9 percentage points to 9.93–15.25%.

Future exploration of the outer planets requires advanced nuclear power systems capable of providing electric power from a few Watts to hundreds of kilowatts for 7–10 years. For these planets, solar power is not an enabling option due to the progressively weaker solar brightness; on Mars, it is ∼45% of that in earth orbit, <4% on Jupiter and essentially nil on other planets farther out. The solar option has been considered for a number of robotic and spacecraft missions to Mars. These missions typically have limited scope and duration from a few days to several months and require only a few to 10’s of Watts of electrical power. For higher power missions to either Mars or the farthest planets, such as Jupiter, Saturn and Pluto, the solar option is not a viable one. Future exploration of these planets will require developing advanced energy conversion technologies that could be used with either a radioactive or a nuclear reactor heat source to provide a wide range of electrical power levels for 7–10 year missions, or even longer. These nuclear power systems operate continuously and independently of the sun. Space reactor power systems (SRPSs) and advanced radioisotope power systems (ARPSs) each enable certain classes of missions. The SRPSs use fission nuclear reactors capable of generating hundred to thousands of kilowatts of thermal power continuously for 7–10 years. These reactors do not start until the spacecraft is fully and safely deployed into space. ARPSs use SOA general purpose heat source (GPHS) bricks that have been used in radioisotope thermoelectric generators (RTGs) on numerous planetary missions for more than three decades [1], [2] and [3]. Each GPHS brick (or module) is loaded with four 238PuO2 fuel pellets that each generates ∼62.5 W of thermal power by the radioactive decay of the 238Pu isotope. The relatively long half life (87 years) of 238Pu makes it suitable for missions of 5 or more years with a small decrease in the thermal power to the end-of-mission (EOM). However, because of the low thermal power density of the 238PuO2 fuel (∼0.4 kWth/kg) and its high density (>1000 kg/m3), limited availability and high cost, ARPSs are only practical for those missions requiring a few tens to a thousand of Watts of electricity. Therefore, ARPSs are enabling deep space and long duration surface and limited subsurface exploration missions on Mars and the farthest planets in the solar system. On the other hand, SRPSs are enabling high power interplanetary missions to propel large spacecraft using a multitude of high power electric propulsion engines requiring 100–250 kWe, and even more. SRPSs markedly shorten the travel time to destination planets, significantly increase the delivered payload mass and provide ample electrical power for the science payload, fast communication and high data transmission rates, surface and subsurface operations and future space colonies. Unlike ARPSs, SRPSs could be designed to operate at variable power levels and for multiple shutdowns and restarts, as needed. A key to enhancing the performance of both ARPSs and SRPSs and reducing their masses and mission costs is developing high efficiency, energy conversion technologies. For ARPSs, candidate technologies are those that could increase the conversion efficiency of the system by >50%, compared to SOA RTGs (∼4.6%), and decrease the system’s specific mass in kg/kWe (or Alpha) by a similar percentage. It is worth noting that the projected Alpha for ARPSs (∼100–200 kg/kWe) is typically 4–5 times that for SRPSs, which explains why the latter are the preferred choice for high electrical power missions. SOA RTGs use SiGe thermoelectric converters that operate typically at hot and cold junctions temperatures of 1273 and ∼566 K, respectively. These converters have accumulated tens of years of remarkable performance on many missions to the sun and various planets in the solar system, such as Mars, Jupiter and Saturn [1], [2] and [3]. SiGe thermoelectric converters have also been used, or considered for use, in SRPSs with a system efficiency of up to 4.6% [4] and [5]. A SOA RTG typically generates ≃290 We, however, its low efficiency (<5%) increases the amount of 238PuO2 fuel needed. Earlier radioisotope space power systems had used lead telluride (PbTe) and silver antimony germanium telluride (TAGS-85) thermoelectric converters at hot and cold junction temperatures of ∼785–800 and 450 K, respectively, and system efficiencies of ∼6.2–6.4% [1], [2] and [6]. The thermoelectric efficiency is directly proportional to the average figure of merit, Z of the thermoelectric materials in the n- and p-legs, where Z = (α2/ρk). As shown in Fig. 1, each thermoelectric material has high Z values only in a certain temperature range. Thus, in segmented thermoelectric converters (STCs) [7], [8], [9], [10], [11], [12], [13] and [14], the p- and n-legs each may be comprised of two or more segments of different but compatible TE materials, depending on the value of the cold junction temperature. The length of each segment is sized to operate in the temperature range in which the thermoelectric material has the highest Z value. The segments in the n- and p-legs are joined with minimal thermal and electrical interface resistances and no or minimal materials diffusion across the interfaces. STCs comprised of BiTe cold segments and Skutterudites [15] and [16] hot segments could be used in waste heat recovery in heavy trucks and a host of other terrestrial applications in which the hot junction temperature is 673–873 K and the cold junction temperature is 298 K. The conversion efficiencies of these STCs could be 13–15% and 10.9–13.3%, when assuming zero side heat losses and a contact resistance per leg of 150 μΩ cm2 and zero, respectively [8], [9] and [10]. Full-size image (31 K) Fig. 1. FOMs of different thermoelectric materials. Figure options Skutterudites with high Z T from ∼0.92 to 1.48 in the temperature range from 300 to 973 K have recently been developed at the Jet Propulsion Laboratory (JPL) in Pasadena, California, [15] and [16]. A number of STCs unicouples have been fabricated using p-type CeFe4CoSb12 and Bi2Te3-based alloys and n-type CoSb3 based alloy and tested at cold and hot junction temperatures of 300 and 973 K, respectively [7], [12] and [13]. Recent tests performed both at JPL and the University of New Mexico’s Institute for Space and Nuclear Power Studies (UNM-ISNPS) demonstrated peak conversion efficiency for these STCs of 13.8%. However, since SiGe is not compatible with Skutterudites and because no other materials have yet been identified for operating at >973 K, CTCs could take advantage of the high temperature FOM of SiGe and the high FOM of Skutterudites for temperatures ⩽973 K [14] and [17] ( Fig. 1). The CTCs optimized to operate at hot junction temperatures up to 1273 K could have higher conversion efficiency than SOA SiGe, reducing the amount of 238PuO2 fuel, the ARPS total mass and the mission cost [17]. This paper presents the performance results of three CTCs for use in ARPSs at significantly higher conversion efficiencies than that of SOA RTGs. The CTCs investigated are optimized for maximum efficiency operation. They have top SiGe unicouples and bottom unicouples with one of the following compositions (Fig. 1 and Fig. 2): (a) TAGS-85 (p-leg) and 2N–PbTe (n-leg); (b) p-leg of CeFe3.5Co0.5Sb12 and n-leg of CoSb3; and (c) segmented p-leg of CeFe3.5Co0.5Sb12 and Zn4Sb3 and n-leg of CoSb3. The performance analyses of the CTCs are conducted for a nominal hot junction temperature for the SiGe top unicouples of 1273 K, three input thermal power values, Qin,CTC, of 1, 2 and 3 Wth and three heat rejection temperatures, TR, from the bottom unicouples of 375, 475 and 575 K. The values of Qin,CTC cover a wide range of heat fluxes in the n- and p-legs that are inclusive of potential integration into either ARPSs or SRPSs. The top and bottom unicouples in the CTCs ( Figs. 2(a)–(c)) are thermally, but not electrically, coupled. Also, the rates of heat rejection from the top SiGe unicouples are the same as the rates of heat input to the bottom unicouples. The analyses assume zero side heat losses and a contact resistance of 150 μΩ cm2 per leg in both the top and bottom unicouples; however, the effect of reducing this resistance to a more realistic value of 50 μΩ cm2 per leg on the performance of the CTCs is investigated. Full-size image (38 K) Fig. 2. Schematics of cascaded thermoelectric converters (CTCs). (a) CTC with TAGS-85/2N–PeTe bottom unicouple. (b) Segmented C with unicouple. (c) CTC with Skutterudites unicouple.

The performance of three CTCs is investigated for potential use in ARPSs having electric power requirements ranging from 20.0 to 1.0 kWe to enable a number of future space exploration missions. The CTCs have significantly higher conversion efficiencies than SiGe convertors used in SOA RTG. Each is comprised of a SiGe top unicouple that is thermally, but not electrically, coupled to a bottom unicouple with one of the following thermoelectric materials: (a) p-leg of TAGS-85 and n-leg of 2N–PbTe; (b) p-leg of CeFe3.5Co0.5Sb12 and n-leg of CoSb3; and (c) segmented p-leg of CeFe3.5Co0.5Sb12 and Zn4Sb3 and n-leg of CoSb3. The analyses are performed at constant hot shoe temperature = 1273 K for the top SiGe unicouples, which is the same as that of the SiGe unicouples in SOA RTG, but the values of the cold junction temperature depend on the thermoelectric materials for the bottom unicouple (780 K for the CTC in Fig. 2(a) and 980 K for the CTCs in Figs. 2(b) and (c)). The cross section areas of the n- and p-legs in the top SiGe unicouples are first optimized at the nominal hot and cold junction temperatures and thermal power input of 1.0, 2.0 or 3.0 Wth. Then, the dimensions of the n- and p-legs of the bottom unicouples in the CTCs are optimized for maximum efficiency operation, subject to the following constraints: (a) constant hot junction temperatures that are 20 K lower than the cold junction temperature of the top SiGe unicouples; (b) the thermal power input is the same as the rejection thermal power from the top SiGe unicouple; (c) nominal cold junction temperatures of 400, 500, or 600 K. The heat rejection temperatures are taken 25 K lower than the cold junction temperatures of the bottom unicouples. For the CTCs, the lengths of the top and bottom unicouples are equal (10 mm) and the nominal operation parameters (conversion efficiency and electrical power) are those corresponding to the peak conversion efficiency of the top SiGe unicouples. When assuming a thermal efficiency of 90% and electric losses of 5% for an ARPS with CTCs, its conversion efficiecy could be 40–110% higher than that of SOA RTG, depending on the value of the heat rejection temperature. Such high ARPS’s efficiencies would significantly reduce the amount of 238PuO2 fuel, the total system mass and the mission cost. The calculated nominal conversion efficiencies and electrical powers of the CTCs with a contact resistance of 150 μΩ cm2 per leg in the top and bottom unicouples are conservative. Decreasing this resistance to a more realistic 50 μΩ cm2 increases the estimates for the ARPS nominal efficiency and electrical power by ∼0.3–0.9 percentage points, depending on the thermoelectric materials of the bottom unicouples and the heat rejection temperature. At the high radiator temperature of 575 K, the CTCs could be used in space nuclear reactor power systems for electrical powers of 10–30 kWe to enable future planetary surface missions without requiring a very large radiator. For higher electrical power levels, however, higher radiator temperatures of 650 to 750 K would be needed for a manageable radiator area and attractive system specific power and total mass and size. The SiGe used in the top unicouples of the three CTCs analyzed in this paper has an excellent track record on many missions that employed RTGs. One of the CTCs developed and investigated in this paper uses a bottom unicouple having a p-leg of TAGS-85 and an n-leg of 2N–PbTe (Fig. 2(a)). These thermoelectric materials have well known properties and fabrication and handling techniques and space flight experiences, but less than SiGe. The other two CTCs in Figs. 2(b) and (c) employ segmented and Skutterudites bottom unicouples. Skutterudites have high FOMs in the temperature range from ∼450 to 973 K and demonstrated efficiency of ∼13.8% at hot and cold junction temperatures of 973 and 300 K, however the device technology is currently at TRL-3 and could be advanced to TRL-5 within the next 2–3 years. These materials and both TAGS-85 and 2N–PbTe used in the bottom unicouples of the CTCs in Fig. 2 are heavier than SiGe, but the high efficiencies of the CTCs more than compensate for the high densities of these materials, resulting in higher specific power ARPSs and much less 238PuO2 fuel than SOA RTGs.