PARAMETRIC ANALYSES OF VAPOR-ANODE, MULTITUBE AMTEC CELLS FOR PLUTO/EXPRESS MISSION Mohamed S. El–Genk and Jean–Michel Tournier Institute for Space and Nuclear Power Studies/Dept. of Chemical and Nuclear Engineering School of Engineering, The University of New Mexico Albuquerque, NM 87131 (505) 277 – 5442 / FAX: – 2814, E-Mail:
[email protected] Abstract A detailed AMTEC Performance and Evaluation Analysis Model (APEAM), developed to predict the performance of next–generation, Pluto/Express multitube cells, was used to evaluate the effects of various design changes on the cell performance. These changes were: using a CREARE condenser; changing the number of BASE tubes and the electrode length; using other electrode materials; using molybdenum as the structural material on the hot side of the cell; and using highly reflective rhodium coatings in the low vapor pressure cavity of the cell. The present analyses utilized a PX-5A type cell, with a CREARE condenser, a molybdenum (Mo) circumferential radiation shield, and 7 BASE tubes with 29 mm-long electrodes. Analyses performed for an output load voltage of 3.5 V per cell have shown that: (a) A stainless steel (SS) cell with improved electrodes (50% lower contact resistance than TiN, and an increased exchange current, B = 200 A.K1/2/Pa.m2) could deliver 7.1 We with an efficiency of 19.5; (b) When Mo was substituted for the SS on the hot side of this cell, the electrical power output increased to 8.8 We and the cell efficiency increased to 20.2%; (c) Using rhodium coatings in the Mo/SS cell increased the electrical power output to 9.7 We, and the peak conversion efficiency of the cell by 2.3 points, to 22.5%. INTRODUCTION NASA’s 12-year Pluto/Express Mission is considering using Alkali–Metal Thermal–to–Electric Conversion cells (AMTEC), in conjunction with a General Purpose Heat Source (GPHS), to supply a total system electric power output of 130 We (Mondt et al. 1997). The high efficiency (15%–25%) and high power density of AMTEC, at a relatively high heat–rejection temperature (about 600 K) and low heat source temperature (below 1200 K), make them an attractive option for this mission, and future space applications. Earlier planetary exploration missions, such as Pioneer, Voyager and Galileo, have used thermoelectric (TE) energy conversion. The low conversion efficiency of TEs (5–7 %) required using larger and heavier GPHSs. AMTEC cells for the Pluto/Express spacecraft would be 3 to 4 times more efficient, requiring less Pu238 for the GPHS, hence reducing the power system’s mass by about 60%. A comprehensive modeling and testing program is underway at the Air Force Research Laboratory’s (AFRL) Power Systems Branch, jointly with the University of New Mexico’s Institute for Space and Nuclear Power Studies (UNM-ISNPS). The objective of this program is to demonstrate the readiness of AMTEC cells technology for flight on the NASA Pluto/Express and other future missions. Schock et al. (1997) have proposed a 28-volt power system for the Pluto/Express spacecraft, which uses 16 multitube AMTEC cells connected in a 8 x 2 series-parallel network, and 2 or 3 GPHS modules, for a total thermal power of 440 W or 660 W, respectively. To fulfill the mission power needs of 130 We, each of the system AMTEC cells would have to produce above 8 We of electrical power output at an average load voltage of 3.5 V, and a conversion efficiency greater than 20%. To date, PX-series cells, fabricated by Advanced Modular Power Systems, and tested at AFRL at hot side temperatures near 1200 K, have delivered peak electrical power outputs up to 5 We per cell, with conversion efficiencies near 15% (Tournier and El-Genk 1998c). The objectives of this work were to perform parametric analyses, evaluate several design options and demonstrate the potential of PX-series cells for achieving the program’s specific goals of: (a) conversion efficiencies in excess of 20%, at typical hot side and cold side temperatures of 1200 K and 623 K, respectively; and (b) an electric power ouput above 8 We, per cell, at a load voltage of 3.5 V. The parametric analyses of the PX-series cells were performed using an integrated AMTEC Performance and Evaluation Analysis Model (APEAM). This model is
being developed as part of the AFRL/UNM-ISNPS joint program. PX-SERIES, VAPOR-ANODE, MULTITUBE AMTEC CELLS The PX–series cells designed by Advanced Modular Power Systems (AMPS), and tested in vacuum at AFRL, had 5, 6 or 7 BASE tubes each, brazed to a stainless steel support plate, and a central felt–metal wick for returning the liquid sodium working fluid to the cell evaporator. The reference cell used in the present analyses was PX-5A. A cross-sectional schematic of the PX-5A cell is shown in Figure 1. The low–pressure side (cathode) and high– pressure side (anode) of the β”-Alumina Solid Electrolyte (BASE) were covered with a TiN porous electrode and molybdenum mesh collector, and the BASE tubes were electrically connected in series. The heat input to the cell hot plate was transported by conduction and radiation to the BASE tubes and the evaporator structure. This PX– series cell had a circumferential radiation shield, placed in the cell cavity above the BASE tubes, to reduce parasitic heat losses to the cell wall, and a solid conduction stud between the cell hot end and the BASE tubes support plate (Figure 1). This cell also used a conical evaporator, which provided a larger surface area for evaporating the liquid sodium returning from the condenser. More details on the operation principle of AMTEC cells can be found in Tournier et al. (1997). Condenser (cold end)
Qair
Liquid-return artery (wick) Cell wall
thermal shield
Qloss
Low-Pressure Sodium Vapor
Evaporator Standoff BASE tubes (with Electrodes) BASE tubes support plate
Hot Plenum (High-Pressure Sodium Vapor)
Cell hot end
Qin
stud
FIGURE 1. A Schematic of PX-5A Vapor-Anode, Multitube AMTEC Cell (Not to Scale). MODEL DESCRIPTION APEAM was developed in Standard Fortran 77, and runs on a PC Compatible DOS machine. This model consists of three major building blocks, which are interactively coupled: (a) a vapor pressure loss model, which simulates free–molecule, transition, and continuum vapor flow regimes in the low pressure cavity of the cell, and calculates the vapor flow pressure losses on the cathode side, as well as the sodium pressures at the interfaces between the electrodes and BASE (Tournier and El-Genk 1998a); (b) a cell electrochemical and electrical model, which determines the internal polarization and concentration losses, the electrical resistances of the BASE and of the current collectors, and calculates the cell total electric current (Tournier et al. 1997); the electrical model allows the current density to vary axially along the BASE tubes; and (c) a radiation/conduction heat transfer model, which calculates internal parasitic heat losses and temperatures within the high and low vapor pressure cavities of the AMTEC cell (Tournier and El-Genk 1998b). The radiation/conduction model incorporated the effects of having a circumferential heat shield above the BASE tubes, and a conduction stud between the hot plate
of the cell and the BASE tubes’s support plate. An efficient iterative solution procedure was developed to couple the three models above to calculate the cell operation parameters (Tournier and El-Genk 1998c). The integrated AMTEC model (APEAM) is described in details in three companion papers in the proceedings (Tournier and ElGenk 1998a, 1998b and 1998c), and in Tournier et al. (1997). BENCHMARK OF MODEL RESULTS WITH EXPERIMENTAL DATA APEAM was benchmarked by comparing its predictions of the operation parameters with measurements for PX4C, PX-5A and PX-3A cells, that were tested in vacuum at AFRL (Merrill et al. 1997). These AMTEC cells had 25.4 mm-long TiN electrodes, covered with molybdenum current collectors. The temperature–independent exchange current and the contact resistance between the TiN electrodes and BASE were taken equal to 120 A.K1/2/Pa.m2 and 0.08 Ω.cm2, respectively (Tournier and El-Genk 1998c). APEAM’s predictions of the performance parameters of PX-4C, PX-5A and PX-3A cells were in good agreement with experimental data (Tournier and El-Genk 1998c). These parameters were the load electrical power output, the cell I-V characteristic, the hot and cold end temperatures of the cell, and the condenser heat rejection. Such good agreement confirmed the soundness of the modeling approach and the solution procedure used. Results showed that the use of a CREARE condenser in PX-5A, in lieu of the mesh pad condenser of PX-4C, reduced the parasitic heat losses in the cell, and increased the cell electrical power output and the predicted conversion efficiency. Best performance was obtained for PX-3A at a hot side temperature of 1173 K. This cell reached a peak electrical power of 4.6 We and a predicted peak conversion efficiency of 14.2 %, in the experiment. PARAMETRIC ANALYSIS AND DESIGN RECOMMENDATIONS OF AMTEC CELLS In this paper, the effects of various design changes, on improving the electrical power output and conversion efficiency of the PX-series, vapor-anode multitube AMTEC cell, were investigated. The present analysis evaluated the effects on the PX-5A cell performance of: (a) reducing the heat losses through the cell wall; (b) using a CREARE condenser; (c) changing the number of BASE tubes and the electrode length; (d) changing the electrode material; (e) using molybdenum as a structural material on the hot side of the cell; and (f) using highly reflective rhodium coatings in the low vapor pressure cavity of the cell. The geometry of the reference cell used in this study was identical to that of PX-5A (Figure 1), except that the evaporator wick surface was flat, and located 16 mm above the BASE tubes support plate. The thickness of the evaporator standoff was adjusted to ensure that the temperature margin, ∆T, in the cell (the temperature difference between the cold end of the BASE tubes and the evaporator) was equal to 50 K. Such a positive temperature margin prevents condensation of sodium vapor inside the BASE tubes, hence avoiding electrical shortage in the cell. Finally, the analysis was performed at constant fixed hot end and cold end temperatures. The hot side temperature was 1200 K, to avoid overheating the BASE tubes brazes in the cell. The cell condenser temperature was 623 K, which is near optimum for AMTEC operation, and is also compatible with the heat rejection of the proposed PX-power system (Schock et al. 1997). The various design changes investigated in the present analyses are listed in Table 1. This table also shows the predicted performance parameters of the cells at the three conditions of interest: (a) at the peak electrical power output; (b) at the peak conversion efficiency; and (c) at a load voltage of 3.5 V. Effect of Reducing Heat Losses through the Cell Wall In the experimental setup at AFRL, the side wall of the AMTEC cell was insulated using 8 cm-thick Min-K (Merrill et al. 1997). The cell wall was thermally coupled to the surrounding Min-K through a radiation gap, and heat was dissipated at the insulation surface by radiation to the vacuum chamber. Figure 2a shows that about 8 W were lost through the cell wall, through the Min-K insulation. These heat losses were almost independent of the load resistance (or the cell electrical current), and amounted to about 20% of the heat input to the cell, when the load resistance RL = 0.8 Ω. The cell provided a peak electrical power of 5.0 We, at a conversion efficiency of 12.4% (Figure 2b). When the cell wall was assumed adiabatic, the efficiency of the cell increased by 2.9 points, to 15.3% (Table 1, cases #1 and 3). Although an adiabatic wall is not a realistic boundary condition, the multitube AMTEC cells in a GPHS AMTEC generator would experience less heat losses through their side wall, than in the vacuum tests at AFRL. Therefore, the performance of the AMTEC cells in the PX-power system would be
expected to be better than measured in vacuum test at AFRL. 20
Wall Heat Losses (W)
(a)
Conversion Efficiency, η (%)
10 Min-K Insulation (AFRL Test Setup)
8 6 4 Multifoil (GPHS-AMTEC Generator) 2 Adiabatic Cell Wall 0
0
2
4
6
8
Adiabatic cell wall 15
10 Min-K insulation (AFRL test setup) 5 Multifoil (GPHSAMTEC generator) 0 0
10
(b)
2
External Load Resistance, RL (Ω)
4
6
8
10
External Load Resistance, RL (Ω)
FIGURE 2. Cell Wall Heat Losses and Conversion Efficiency (Cases #1, 2 and 3). In an attempt to perform a conservative, but more realistic assessment of the performance of AMTEC cells for a PX-power system, the heat losses through the cell wall in the following analyses were kept constant at 2 W. This case was labelled “multifoil insulation” in Figure 2a. The reduction in wall heat losses from 8 W to about 2 W caused a 2.2 points increase in peak cell efficiency, from 12.9% to 15.1% (Table 1, cases #1 and 2). In the following analyses, the “multifoil insulation” case was selected as the base case. 20 (a) 5
Conversion Efficiency, η (%)
Electrical Power, Pe (We)
6
CREARE condenser (ε = 0.05)
4 3 2 Mesh pad condenser (ε = 0.15) 1 0
2
4
6
8
External Load Resistance, RL (Ω)
10
CREARE condenser (ε = 0.05)
(b) 15
10
5 Mesh pad condenser (ε = 0.15) 0 0
2
4
6
8
10
External Load Resistance, RL (Ω)
FIGURE 3. Effect of Condenser Type on Cell Electrical Power and Conversion Efficiency (Cases #2 and 4). Effect of Condenser Design The effect of changing the condenser type on the cell performance is shown in Figure 3. The stainless steel mesh pad covering the condenser surface was replaced by a CREARE type condenser (Crowley and Izenson 1993). The latter was designed to ensure the formation of a continuous film of liquid sodium on its surface. The effective emissivity of the CREARE condenser surface was taken equal to 0.05 in the model, while that of the SS mesh pad condenser was assumed equal to 0.15. As shown in Figure 3, the CREARE condenser reduced the parasitic radiation losses to the condenser, thus increasing the BASE tubes and evaporator temperatures. The cell peak electrical power increased by 0.5 We, from 4.9 We to 5.4 We, and the peak cell efficiency increased by 2 points, from 13.1% to 15.1% (Table 1, cases #2 and 4). The use of a CREARE condenser is retained, in the analyses. Effect of Changing the Number of BASE Tubes and the Length of the Electrodes The effect of changing the number of BASE tubes in the cell is investigated in this section. As shown in Figure 4a, the cell internal resistance increased linearly with increasing number of BASE tubes, since the BASE tubes
were connected in series, electrically. The internal electrical resistance of the cell also decreased, inversely proportionally to increasing the electrode length. The major components of the cell internal resistance were: (a) the ionic resistance of the BASE, which is proportional to the β”-alumina solid electrolyte thickness; and (b) the contact resistances between the BASE and the metallic electrodes, and between the electrodes and the current collectors. The ohmic losses in the 40-mesh, molybdenum current collectors were negligible. Figure 4c shows that the cell electrical power output increased with increasing number of BASE tubes. Near the peak electrical power, every BASE tube in the cells provided about 0.9 We (Table 1, cases #5, 2, 6 and 7). However, the electrical power output per tube decreased slowly, as the number of BASE tubes in the cell increased. The increase in cell voltage, and the associated increase in the cell electrical current, caused the charge-exchange polarization losses, and concentration losses, also to increase, since the sodium vapor pressure at the BASE/cathode interface increased with sodium mass flow rate, or cell current (Figure 4b). Also, the larger the number of BASE tubes in the cell, the larger the radiation view factor between the BASE tubes bundle and the colder surfaces in the cell cavity, above the BASE tubes. As a result, increasing the number of BASE tubes increased the internal parasitic losses in the cell, reducing somewhat the temperatures of the BASE tubes and the evaporator. 7
0.38 0.36
7 β-tubes, LE= 29.0 mm
0.34
6 β-tubes, LE= 25.4 mm
0.32 0.30
5 β-tubes, LE= 25.4 mm
0.28 0.26 0
Na Pressure at BASE/Cathode (Pa)
Electrical Power, Pe (We)
7 β-tubes, LE= 25.4 mm
(a)
2
4
6
8
50
7 β-tubes 5 6 β-tubes
4 3 2 1
5 β-tubes 2
4
6
8
10
20 (b)
40 7 β-tubes, LE = 29 mm 30 7 β-tubes 6 β-tubes 20 5 β-tubes 10
(c)
7 β-tubes, LE= 29 mm
6
0 0
10
Conversion Efficiency, η (%)
Cell Internal Resistance, Rint (Ω)
0.40
0
2
4
6
8
External Load Resistance, RL (Ω)
10
(d)
7 β-tubes, LE= 29 mm
15
10 7 β-tubes 5
6 β-tubes 5 β-tubes
0 0
2
4
6
8
10
External Load Resistance, RL (Ω)
FIGURE 4. Effects of Changing the Number of BASE Tubes and the Electrode Length, LE, on Cell Performance (Cases #5, 2, 6 and 7). Nonetheless, Figure 4d shows that the overall conversion efficiency of the cell still increased with increasing number of BASE tubes. This is because the temperature margin in the cell was kept constant in the analyses. As more BASE tubes were added to the cell, the evaporator standoff thickness was adjusted to keep the temperature margin at 50 K. Model results have also shown that if the evaporator standoff geometry was kept unchanged, the cell conversion efficiency would have been highest with 6 BASE tubes. However, an identical cell with 7 BASE tubes, covered with longer electrodes, would have delivered 1 We more, at the same peak efficiency. When the number of BASE tubes was increased, while keeping ∆T constant at 50 K, the significant gain in the cell electrical
power output more than compensated for the increase in the cell internal parasitic losses, and the conversion efficiency continued to increase. To improve the electrical power output of an AMTEC cell with 7 BASE tubes, the length of the electrodes was extended from 25.4 mm to 29.0 mm. Note that such increase in the electrode length is possible with the current 40 mm-long BASE tubes. The braze section can be reduced in height to less than 7 mm, which was done in the PX3A cell by AMPS. Figure 4a shows that extending the electrode length covering the BASE tubes reduced the cell internal resistance from 0.38 Ω to 0.34 Ω, and increased the cell current (which varied proportionally with the area of electrodes in the cell), increasing the electrical power output of the cell by 0.8 We, from 6 We to 6.8 We. As a result, the peak efficiency of the cell increased to 16.9% (Figure 4d). Note that the peak electrical power output and peak conversion efficiency shifted to a higher load resistance (lower cell currents) as the cell internal electrical losses increased (Figures 4c and 4d). In summary, a cell having 7 BASE tubes with 29.0 mm-long TiN electrodes, can provide 6.8 We, at a conversion efficiency of 15.8%, when operated at a hot end temperature of 1200 K. This is 1.5 We more electrical power and 1.5 points more conversion efficiency than the actual PX-5A cell, which had 6 BASE tubes with 25.4 mm-long TiN electrodes (Table 1, cases #2 and 7). Effect of Changing the Electrode Material AMTEC cells which use TiN electrodes have relatively large internal electrical losses (Tournier et al. 1997). These electrodes are about 90% porous, their temperature-independent exchange current is relatively low (B = 120 A.K1/2/Pa.m2), and the contact resistance between the BASE, electrode, and current collector is relatively large (Rcont = 0.08 Ω.cm2). As a result, the concentration losses in the cell (the effect of sodium vapor pressure at the BASE/cathode interface) are small, compared to the charge-exchange polarization and internal ohmic losses. A cell with seven 0.5 mm-thick BASE tubes, and 29.0 mm-long TiN electrodes, would have an internal resistance of 0.34 Ω (Figure 4a). The ionic resistance of the BASE contributes 0.16 Ω (or 0.14 Ω.cm2), while the contact resistances with the electrodes amount to 0.18 Ω (2 x 0.08 = 0.16 Ω.cm2), or about 53% of the total internal resistance. 20
Electrical Power, Pe (We)
(a)
Conversion Efficiency, η (%)
10 2
Rcont= 0.04 Ω.cm , B = 200
8
2
Rcont= 0.04 Ω.cm , B = 120
6 4 2
2
Rcont= 0.08 Ω.cm , B = 120 0 0
2
4
6
8
External Load Resistance, RL (Ω)
10
(b)
2
Rcont= 0.04 Ω.cm , B = 200
15
10 2
Rcont= 0.04 Ω.cm , B = 120 5
2
Rcont= 0.08 Ω.cm , B = 120 0 0
2
4
6
8
10
External Load Resistance, RL (Ω)
FIGURE 5. Effect of Electrode Material on Cell Performance (Cases #7, 8 and 9). Figure 5 shows the effect of decreasing the contact resistance on the performance parameters of the AMTEC cell. A 50% reduction in the contact resistance (from 0.08 Ω.cm2 to 0.04 Ω.cm2 ) reduced the cell internal resistance by 26% (from 0.34 Ω to 0.25 Ω), resulting in a 0.7 We increase in the cell peak electrical power (to 7.5 We) and a 0.8 point increase in the cell peak efficiency (to 17.7%). Furthermore, an increase in the temperature-independent exchange current, from B = 120 to B = 200 A.K1/2/Pa.m2, reduced the polarization losses in the cell, increasing the peak electrical power of the cell by another 0.8 We (to 8.3 We), and the cell peak efficiency by another 0.9 point, to about 18.6% (Table 1, cases #8 and 9).
In summary, the development of advanced electrode materials, with lower interfacial contact resistance and higher exchange current, would improve the AMTEC cell performance. A reduction of Rcont by 50%, to 0.04 Ω.cm2, or increasing the exchange current from 120 to 200 A.K1/2/Pa.m2, are only moderate changes, that could be achievable in the near term (2 to 3 years). For example, oxide-free Mo electrodes have contact resistances < 0.015 Ω.cm2, and B > 400 A.K1/2/Pa.m2 (Sievers and Bankston 1988). Unfortunately, Mo electrodes degrade quickly when operated at high temperatures. Also, exchange currents as high as 200 A.K1/2/Pa.m2 have been measured for uncontaminated Rh2W electrodes at Jet Propulsion Laboratory (Ryan et al. 1992). In addition to demonstrating high performance, the electrodes must also demonstrate stability of operation at high temperature, for long periods of time, comparable with the expected mission lifetime (10 to 15 years). Effect of Changing the Hot-Side Structural Material and Using High-Reflectivity Coatings Analyses of the heat transfer in the multitube AMTEC cell have shown that the heat was transported from the cell hot end to the BASE tubes and the evaporator standoff mostly by conduction (Tournier and El-Genk 1998b). Therefore, substituting the highly thermally conductive molybdenum for stainless steel on the hot side of the cell (plenum, conduction stud, support plate, the portion of the wall facing the BASE tubes, and evaporator standoff) would increase the structure’s conductance on the hot side of the cell by a factor 3. The added benefit of using Mo is its much lower emissivity (ε = 0.08, compared to ε = 0.25–0.30 for SS), which would reduce the parasitic radiation heat losses from the hot support plate, evaporator standoff, and hot cell wall, to the cooler surfaces of the cell cavity, above the BASE tubes. Figure 6 shows the results of substituting Mo for SS, as the structural material on the hot side. The cell peak electrical power increased by 1.4 We (to 8.2 We), while the peak efficiency increased by 0.5 points (to 17.5%). The BASE tubes and evaporator temperatures, as well as the parasitic heat losses in the cell, increased. The BASE tubes brazes temperature increased by about 10 K, while the BASE tubes cold end temperature increased by about 20 K. Note that it was necessary to reduce the thickness of the evaporator standoff rings in order to keep a 50 K temperature margin in the cell. Model results have shown that if the evaporator standoff geometry was kept unchanged, the substitution of Mo for SS on the hot side would have resulted in a larger gain in cell electrical power output; however, the temperature margin would become negative, ∆T = -20 K. Therefore, the cross-section area of the evaporator standoff rings would have to be reduced, in order to maintain an adequate temperature margin in the Mo/SS cell. In summary, the use of Mo on the hot side of the AMTEC cell resulted in significant improvements in the cell performance. This design change requires that the Mo hot wall and evaporator standoff be brazed to a lower thermal conductivity material, such as stainless steel, which is a technological challenge. 20 (a)
8
Conversion Efficiency, η (%)
Electrical Power, Pe (We)
9 Mo/SS cell, Rh coatings
7 6
Mo/SS cell
5 4 3
SS/SS cell
2 1 0
2
4
6
8
External Load Resistance, RL (Ω)
10
Mo/SS cell, Rh coatings
(b)
15 Mo/SS cell 10
SS/SS cell
5
0 0
2
4
6
8
10
External Load Resistance, RL (Ω)
FIGURE 6. Effect of Changing Structure Material and Using High-Reflectivity Coatings on Cell Performance (Cases #7, 10 and 11). When the artery, the internal circumferential heat shield, and the cell wall above the BASE tubes were coated with a high-reflectivity material, such as rhodium (ε = 0.05–0.07), the parasitic heat losses in the cell were reduced and the BASE tubes and evaporator temperatures increased, increasing both the cell electrical power output and
conversion efficiency. Figure 6a shows that the use of rhodium coatings increased the cell peak electrical power by 0.7 We (from 8.2 We to 8.9 We), and increased the cell peak conversion efficiency (Figure 6b) by 2.1 points (Table 1, cases #10 and 11). The Mo/SS cell with Rh coatings could deliver 7.8 We at the peak efficiency of 19.6%, at a load voltage of 3.5 V. Note that the performance of this cell narrowly missed the goals of 8 We output with conversion efficiencies above 20%. In conclusion, high-reflectivity coatings could improve the performance of the AMTEC cell, given that these coatings can survive at high temperature for the entire time of the mission, in a sodium vapor environment. Improved Performance Parameters of Adiabatic PX-Series AMTEC Cells In order to evaluate the maximum performance potential of PX-series AMTEC cells, the analysis was pursued, assuming an adiabatic cell wall, and incorporating the design changes investigated earlier. Figure 7 shows that a stainless steel cell with 7 BASE tubes and 29.0 mm-long electrodes (assuming 50% lower contact resistance than TiN, and a higher exchange current, B = 200 A.K1/2/Pa.m2) had a peak electrical power of 8.5 We and a peak conversion efficiency of 19.6%, when operated at a hot end temperature of 1200 K. This cell could deliver 7.1 We at an efficiency of 19.5% (close to the peak value), for a load voltage of 3.5 V. When Mo was substituted for SS on the hot side of the cell, the peak electrical power and peak conversion efficiency increased by 2 We and 0.6 point, to 10.4 We and 20.21%, respectively (Table 1, cases #13 and 14). This Mo/SS cell delivered 8.8 We with an efficiency of 20.18% (very close to the peak value), at the same load voltage of 3.5 V. These results represent a 24% increase in the cell electrical power output, and 3.6% increase in the cell conversion efficiency. 25
12 2
Electrical Power, Pe (We)
10
Conversion Efficiency, η (%)
(a)
Mo/SS/Rh (0.04 Ω.cm , B = 200) 2
Mo/SS (0.04 Ω.cm , B = 200)
8
2
SS/SS (0.04 Ω.cm , B = 200)
6 4 2
2
SS/SS (0.08 Ω.cm , B = 120)
0 0
2
4
6
8
5 β-Tubes Braze Temperature (K)
External Load Voltage (V)
4 3.5 V
3 2
2
Mo/SS (0.04 Ω.cm , B = 200) 2
SS/SS (0.04 Ω.cm , B = 200)
1
2
SS/SS (0.08 Ω.cm , B = 120)
0
0
1
2
External Load Resistance, RL (Ω)
FIGURE 7. 15).
15 10
2
Mo/SS (0.04 Ω.cm , B = 200) 2
SS/SS (0.04 Ω.cm , B = 200)
5
2
SS/SS (0.08 Ω.cm , B = 120)
2
4
6
8
10
1200
(b)
2
Mo/SS/Rh (0.04 Ω.cm , B = 200)
Mo/SS/Rh (0.04 Ω.cm , B = 200)
20
0 0
10
2
(c)
3
2
Mo/SS/Rh (0.04 Ω.cm , B = 200)
1190
(d)
2
Mo/SS (0.04 Ω.cm , B = 200)
1180 1170 2
SS/SS (0.08 Ω.cm , B = 120)
1160
2
SS/SS (0.04 Ω.cm , B = 200)
1150
0
2
4
6
8
10
External Load Resistance, R L (Ω)
Improved Performance Parameters of Adiabatic PX-Series AMTEC Cells (Cases #12, 13, 14 and
Figure 7 shows that using rhodium coatings in the Mo/SS cell increased the peak electrical power by 0.8 We (from 10.4 We to 11.2 We), and increased the peak conversion efficiency of the cell by 2.3 points (Table 1, cases #14 and 15). The Mo/SS cell with Rh coatings could deliver 9.7 We with an efficiency of 22.5% (close to the peak
value), when the load voltage = 3.5 V. Using Mo as the structural material on the hot side of the cell, and rhodium coatings on the colder surfaces of the cell, above the BASE tubes, increased the BASE tube brazes temperature (Figure 7d). Substituting Mo for SS on the hot side increased the BASE tube brazes temperature by 12 K, near the peak conversion efficiency, to about 1170 K, while using rhodium coatings only resulted in an additional 3 K increase in the temperature of the brazes. SUMMARY AND CONCLUSIONS In this paper, parametric analyses of PX-type AMTEC cells were performed using the AMTEC Performance and Evaluation Analysis Model (APEAM). APEAM has been benchmarked successfully against experimental data of PX-4C, PX-5A and PX-3A cells, which were tested in vacuum at AFRL. The present analyses evaluated the effects of the following design changes: (a) reducing the heat losses through the cell wall; (b) using a CREARE type condenser; (c) changing the number of BASE tubes and the electrode length; (d) changing the electrode material; (e) using molybdenum as the structural material on the hot side of the cell; and (f) using highly reflective rhodium coatings in the low vapor pressure cavity of the cell. The analyses were performed at fixed hot and cold end temperatures of 1200 K and 623 K, respectively. The analyses employed a PX-5A type cell, with a CREARE condenser, a Mo circumferential radiation shield, and 7 BASE tubes connected in series, with 29 mm-long electrodes. A cell based on this design could deliver a 3.5-V voltage to the load with a conversion efficiency close to the peak value. The peak efficiency always occurred at a load resistance greater than that corresponding to the peak electrical power output, a suitable operating point in the load-following portion of the cell operation. A stainless steel cell, with negligible heat losses through its wall, and TiN electrodes, would deliver 5.9 We at an efficiency of 17.8%, and a load voltage of 3.5 V. Results showed that a stainless steel cell with improved electrodes (50% lower contact resistance, Rcont , than TiN, and a higher exchange current, B = 200 A.K1/2/Pa.m2) could deliver 7.1 We with an efficiency of 19.5%, at a load resistance and voltage of 1.73 Ω and 3.5 V, respectively. Current TiN electrodes have Rcont = 0.08 Ω.cm2, and B = 120 A.K1/2/Pa.m2. When Mo was substituted for SS as the structural material on the hot side of the cell, the cell delivered 8.8 We with an efficiency of 20.2%, at a load resistance of 1.40 Ω and a load voltage of 3.5 V. This design change, however, requires that the Mo hot wall and evaporator standoff be brazed to a lower thermal conductivity material, such as stainless steel, which is a technological challenge. Using rhodium coatings in the Mo/SS cell increased the electrical power output of the cell by 1 We, to 9.7 We, and increased the conversion efficiency of the cell by 2.3 points, to 22.5%, at a load resistance of 1.26 Ω and load voltage of 3.5 V. In conclusion, results of the present analyses showed that Pluto/Express AMTEC cells could achieve performance in excess of the design goals of 8 We at a load voltage of 3.5 V, and a conversion efficiency > 20%, when the different changes investigated herein are incorporated into the cell design. Some of the design changes proposed, however, would require resolving key technology issues, such as brazing of Mo and stainless steel, applying rhodium coating that could last the mission lifetime in a corrosive sodium vapor atmosphere, and developing new electrode materials with lower contact resistance and higher exchange current. Acknowledgments This research is funded by the Power Systems Branch of the Air Force Research Laboratory, Kirtland AFB, Albuquerque, NM, under contract F29601–96–K–0123, to the University of New Mexico’s Institute for Space and Nuclear Power Studies. References
Crowley, C. J., and M. G. Izenson (1993) “Condensation of Sodium on a Micromachined Surface for AMTEC,” in Proc. 10th Symposium on Space Nuclear Power and Propulsion, M. S. El-Genk, ed., American Institute of Physics, NY, NY Conf. Proceeding No. 271, CONF-930103, 2:897-904. Merrill, J., M. J. Schuller, R. K. Sievers, C. A. Borkowski, L. Huang, and M. S. El-Genk (1997) “Vacuum Testing of High-Efficiency Multitube AMTEC Cells,” in Proc. 32nd Intersociety Energy Conversion Engineering Conf., Paper No. 97379, American Chemical Society, 2:1184-1189. Mondt, J. F., M. L. Underwood, and B. J. Nesmith (1997) “Future Radioisotope Power Needs for Missions to the Solar System,” in Proc. 32nd Intersociety Energy Conversion Engineering Conf., Paper No. 97244, American Chemical Society, 1:460-464. Ryan, M. A., B. Jeffries-Nakamura, R. M. Williams, M. L. Underwood, D. O’Connor, and S. Kikkert (1992) “Advances in Materials and Current Collecting Networks for AMTEC Electrodes,” in Proc. 27th Intersociety Energy Conversion Engineering Conf., Paper No. 929007, Society of Automotive Engineers, 3:3.7-3.12. Schock, A., H. Noravian, C. Or, and V. Kumar (1997) “Design and Analysis of Radioisotope Power System Based on Revised Multitube AMTEC Cell Design,” in Proc. Space Technology and Applications International Forum (STAIF-97), M. S. El-Genk, ed., American Institute of Physics, NY, NY Conf. Proceeding No. 387, CONF970115, 3:1411-1423. Sievers, R. K., and C. P. Bankston (1988) “Radioisotope Powered Alkali Metal Thermoelectric Converter Design for Space Systems,” in Proc. 23rd Intersociety Energy Conversion Engineering Conf., Paper No. 889082, Goswami, D. Y., Ed., The American Society of Mechanical Engineers, 3:159-167. Tournier, J.-M., M. S. El-Genk, M. Schuller, and P. Hausgen (1997) “An Analytical Model for Liquid-Anode and Vapor-Anode AMTEC Converters,” in Proc. Space Technology and Applications International Forum (STAIF97), M. S. El-Genk, ed., American Institute of Physics, NY, NY Conf. Proceeding No. 387, CONF-970115, 3:1543-1552. Tournier, J.-M., and M. S. El-Genk (1998a) “Sodium Vapor Flow Regimes and Pressure Losses on Cathode Side of Multitube AMTEC Cell,” in Proc. Space Technology and Applications International Forum (STAIF-98), M. S. El-Genk, ed., American Institute of Physics, NY, NY. Tournier, J.-M., and M. S. El-Genk (1998b) “Radiation/Conduction Model for Multitube AMTEC Cells,” in Proc. Space Technology and Applications International Forum (STAIF-98), M. S. El-Genk, ed., American Institute of Physics, NY, NY. Tournier, J.-M., and M. S. El-Genk (1998c) “AMTEC Performance and Evaluation Analysis Model (APEAM): Comparison with Test Results of PX-4C, PX-5A, and PX-3A Cells,” in Proc. Space Technology and Applications International Forum (STAIF-98), M. S. El-Genk, ed., American Institute of Physics, NY, NY. ---------------------------------------------------------- Nomenclature -------------------------------------------------------------English B LE Mo Pe Q Rcont RL SS
Temperature-independent exchange current (A.K1/2/Pa.m2) Electrode length (m) Molybdenum Cell electrical power output (We) Heat flow (W) Contact resistance between BASE/electrode/ current collector (Ω.cm2) External electrical load resistance (Ω) Stainless steel
T
Temperature (K)
Greek
∆Τ ε η
Temperature margin in cell (temperature difference between cold end of BASE tubes and evaporator) Surface radiative emissivity Cell overall conversion efficiency (%)
Subscript / Superscript
air Air calorimeter at cold end in Input cond Cell condenser loss Side wall losses hot Cell hot end max Maximum value -------------------------------------------------------------------------------------------------------------------------------------------
