NATIONAL AERONAUTICS AND SPACE ADMINISTRATION

Technical Memorandum 33-327

R TG Integration Problem Areas and Parametric Analysis

L. Selwitz

Approved by:

P. Goldsmith, Manager Spacecraft Power Section

JET PROPULSION LABORATORY

CALIFORNIA INSTITUTE OF TECHNOLOGY

PASADENA, CALIFORNIA

February 1, 1967 TECHNICAL MEMORANDUM 33-327

Copyright 0 1967 Jet Propulsion Laboratory California Institute of Technology

Prepared Under Contract No. NAS 7-100 National Aeronautics & Space Administration Contents

1. Introduction ...... 1

II. Interface Constraints ...... 1

111. Generator Oriented Constraints ...... 2

IV . Mission Environment Constraints ...... 2

V . Safety-Oriented Constraints ...... 2

VI . Spacecraft (System)-Oriented Constraints ...... 2

VI1 . Interface Matrix ...... 2

VI11 . Hybrid Systems ...... 5

IX. Parametric Relationships ...... 6

X . Conclusion ...... 9

Appendix . Thermoelectric Definitions ...... 10

1. Thermoelectric Material Effects and Parameters ...... 10

II. Performance Criteria ...... 10

111. Thermoelectric Hardware - Heat Sources ...... 11

IV. Thermoelectric Hardware - Thermoelectric Assemblies ...... 12

V . Thermoelectric Hardware - Heat Rejection ...... 13

VI . Nuclear Applications ...... 13

VII . Mission and Application Considerations ...... 14

Nomenclature ...... 15

Figures

1. Radioisotope thermoelectric generator system relationships ...... 3 2. Hybrid system weight analysis ...... 4

JPl TECHNICAL MEMORANDUM 33321 iii Abstract

Current radioisotope thermoelectric generator (RTG) design concepts for space power systems are often unrealistic or incomplete from two standpoints. They generally lack understanding of, or regard for, the total mission and spacecraft requirements, and are usually unsupported by sufficient reliable data or technical information. Too little differentiation is made between proof-of-principle hard- ware concepts and mission applications. In the former, flights of specific RTGs or families of RTGs are planned for the purpose of determining the feasibility of new RTG sizes, weights, and fueIs; in the Iatter the RTG has functional impor- tance, utilized as a component whose sole purpose is to support the primary goals of the mission.

iv JPL TECHNICAL MEMORANDUM 33-321 RTG Integration Problem Areas and Parametric Analysis

1. Introduction 1966 thermoelectric and fuel capsule state-of-the-art indicates that RTG integration feasibility is indeed real. A practical consideration of radioisotope thermoelec- The extent of the conditions under which this feasibility tric generator power sources for space applications re- exists is, unfortunately, ill-defined, and can be clarified quires an understanding of both the total system into only through a total system approach. Feasibility must which the RTG will be integrated and the interface con- be determined on a mission-to-mission and spacecraft-to- straints imposed by such a system. Stated another way, spacecraft basis. RTG design and integration analysis should be tested against the following standard: Given any variety of mis- sions, with their full spectrum of interface constraints, is II. Interface Constraints the use of an RTG feasible? If so, what RTG size and weight can be expected, all interface constraint condi- Starting with a given mission and only its power re- tions being met? quirements, a first rough-cut RTG sizing is relatively straightforward. Characteristics curves are available for The above premise suggests a wide range of possi- thermoelectric materials, such as lead telluride and sili- bilities : con germanium, and optimum values of thermoelectric module size and weight can be determined for specific (1) An RTG is not feasible for these missions. power and voltage requirements. These temperature (2) An RTG may be feasible for definite and limited dependent values must be based upon the assumed types of missions only. availability of a heat source and heat sink capable of providing a usable thermal flux. Similarly, the thermal (3) RTGs have broad applicability. pattern selected to determine the thermoelectric output (4) RTGs are limited in their potential only to the values can be used for rough determinations of the heat extent that some interface constraints are either source and heat sink sizes and weights. unrealistic or too stringent. However, a mission consists of more than its power The above possibilities, in turn, suggest that an RTG requirements alone. Conditions of long life, severe envi- as an isolated component or piece of hardware is some- ronmental stresses, and Ioad variations, among others, what unrealistic; at no time can the RTG be taken out impose restrictions that strongly affect thermoelectric of context to its ultimate application as part of a total performance. To compensate for anticipated debilitating system. effects that result in RTG performance degradation, or

JPL TECHNICAL MEMORANDUM 33-321 1 even to prevent such effects, if possible, changes in such trix relationship exists between it and the generator areas as generator design, assembly techniques, and op- oriented-constraint area. erating temperature range may have to be made. These a priori changes, when they occur, must be made in light of the overall mission and its interfaces. V. Safety-Oriented Constraints The safety-oriented constraints are externally estab- Four distinct and equally important interface con- lished requirements, compliance with which results in straint areas must be considered in context with the mis- the addition of structural features to the thermoelectric sion power requirements: generator, increasing size and weight but without a com- (1) Generator oriented constraints. mensurate increase in generator output or performance. These external requirements are political, philosophical, (2) Mission environment constraints. and economic, as well as technical in nature, and are (3) Safety-oriented constraints. germane to nuclear-fueled power sources to a degree never experienced with chemical or solar photovoltaic (4) Spacecraft (system)-oriented constraints. devices. Their result is the consideration and use of RTG-spacecraft structural integration members; intact or The mission power requirements may be considered a burn-up re-entry structural and control members; and fifth interface constraint area. thermal and nuclear radiation protection members. The safety-oriented constraints are most traceable, design- wise, to the heat source, but their influence on mission 111. Generator Oriented Constraints science and electronics suggests a direct matrix relation- The generator oriented constraints are those RTG ship with the system-oriented constraint area. operating limits beyond which reliable thermoelectric performance can neither be predicted nor expected. This constraint area provides the primary feedback network VI. Spacecraft (System)-Oriented Constraints necessary for keeping the RTG design consistent with The system-oriented constraints are those restrictions its capabilities. The generator operating limits are essen- imposed upon the generator parameters which, if ex- tially the physical, thermal, and electrical tolerances of ceeded, could be a threat to the achievement of the mis- the thermoelectric module. These, in turn, are influenced sion objectives. That is, the desired generator output by: cannot be achieved with side effects that pose a hazard (1) Ambient thermal conditions. to the spacecraft and the science and electronics it con- Factors to be considered are: (2) Impact and vibration stresses. Radiation tolerance levels of spacecraft compo- (3) Mission time duration. nents and electronics. (4) Ambient atmospheres. Temperature tolerance levels of the spacecraft structure and on-board components.

IV. Mission Environment Constraints Orientation and placement of RTGs with regard for spin or three-axis spacecraft stabilization. The mission environment constraint area differs from the other three constraint areas to the extent that it is Threshold levels of on-board experimentation, such rather inflexible. Once a mission has been selected there as magnetometers and ionization chambers. is relatively little that can be done to change conditions Spacecraft weight allowance. such as Martian or interplanetary ambient atmospheres or temperatures, or launch and entry vibrations and im- Spacecraft and accessory space availability. pacts. The importance of the mission environment con- straints is to point out those areas where generator protection is needed, and to indicate the size and weight Interface Matrix penalties associated with such protection. The fact that Figure 1 illustrates the interrelationship among the the mission environment may impose severe operating requirements, the interface constraints, and the subse- conditions upon the generator suggests that a direct ma- quent factors affecting RTG sizing. It must be realized

2 JPl TECHNlCAl MEMORANDUM 33-327 POWER SOURCE

I I I I GENERATOR I ORIENTED I ORIENTED I I I CONSTRAINTS I CONSTRAINTS I I I + I POWER CONDITIONING RADIATION /LI ON-BOARD .I M SIZE, wt SHIELDING \c1 COMPOfVENTS I HALF-LIFE ATTENUATION TOLERANCE 7 I SPACECRAFT I MODULE - STRU rURE 1 ARTIFACTSSIZE, wt -4 s1z wt I I REQUIREMENTS I ISOTOPE -- 6- --+--- +-- +--.+--+--. ---- tFORM

I . II I I 1 RE-ENTRY I I I INTACT 71 I SAFfTY THERMAL THERMOELECTRIC AMBIENT GENERATOR

II TEMPERATURE I I PHYSICAL RELATIONSHIP LIMIT I 2. THERMAL RELATIONSHIP -- - - 3. RADIOACTIVITY RELATIONSHIP -+- i 1 COMBINATION 2+3 +-- 4 ELECTRICAL RELATIONSHIP ------I

adioisotope thermoelectric generator system relationships

JPL TECHNlCAL MEMORANDUM 33-32 I 3- 4 500 w PEAK POWER

240 -

WEIGHT OF A HYBRID POWER SOURCE SYSTEM CONSISTING - OF AN RTG AND RECHARGEABLE BATTERY PACKAGE AS 200 A FUNCTION OF PEAK POWER, NOMINAL POWER, AND POWER PROFILE-FOR A 24-hr CYCLE BASED ON: RTG SPECIFIC POWER OF 2 w/lb

180 - RTG SIZE BASED UPON FORMULA (G-PN) (7&) + (G) (&AX) = (PMAX)(&AX) WHERE: G=RTG POWER OUTPUT, w TN=NOMINAL POWER DURATION, hr PN = NOMINAL HYBRID POWER &AX=PEAK POWER DURATION, hr PMAXZPEAK HYBRID POWER

300 w PEAK POWER EXAMPLE:300 w PEAK POWER DESIRED FOR 8-hr; 150 w

NOMINAL POWER REQUIRED FOR 16-hr WHAT SIZE RTG IS REQUIRED? WHAT WILL BE THE WEIGHT OF A HYBRID SYSTEM?

ANALYSIS: FROM ABOVE FORMULA, RTG IS SIZED AT 200 w FROM GRAPH, HYBRID SYSTEM WILL WEIGH 140 Ib I. DURING NOMINAL POWER PERIOD, I50w ARE REQUIRED FOR 16-hr, OR 2400w-hr

200 w PEAK POWER 2. 200 w RTG WILL SUPPLY I50 w FOR 16 hr OF NOMINAL POWER (2400 w-hr) AND ALSO PROVIDE 50 wOF BATTERY CHARGING FOR 16-hr (800 w-hr) 3. DURING PEAK POWER PERIOD, 300w ARE REQUIRED FOR 8-hr, OR 2400w-hr I50 w PEAK POWER 4. 200w RTG WILL SUPPLY 200w FOR 8-hr PEAK PERIOD (1600w-hr) AND RECHARGED BATTERY WILL SUPPLY 800w-hr, FOR A TOTAL OF 2400w-hr DURING PEAK PERIOD 5. 200~RTG AT2w/lb WILL WEIGH 1001b.800w-hr BATTERY PACK AT 20w-hr/lb WILL WEIGH 40 Ib 100 w PEAK POWER TOTAL WEIGHT= 100+40 = 140 Ib

50 w PEAK POWER

I I I I I I I I 1 I 1 0 IO 20 30 40 50 60 70 80 90 100 110 120

PERCENT, “NOMINAL POWER“ TO “PEAK POWER: PN/~MAX

Fig. 2.lh.lybrid system weight analysis

4 JPl TECHNICAL MEMORANDUM 33-32 1 y-Q that, for any given mission, RTG feasibility is possible If any of the numbers within the matrix exceed the only when the interrelationships of the interface matrix values assigned to the system or generator-oriented con- are in a state of equilibrium; that is, the ultimate RTG straints, new iterations must be made with respect to the size and weight, after meeting all of the interface con- thermoelectric module, the heat source, or the heat ex- straints, equals the size and weight allowance provided changer design to bring the matrix back to equilibrium. by the mission spacecraft. Thermal analysis provides an excellent example of the Interface equilibrium cannot be achieved by common use of the matrix. Five blocks directly interrelated in this analytical processes; the matrix of all possible design area are: and constraint variations is, for all practical purposes, infinite. RTG feasibility analysis and sizing can be (1) Temperature Range (thermoelectric design). achieved only through the translation of all interrelation- ships into parametric form. (2) Temperature Limitations (generator-oriented con- straint). The starting point for the analysis is thus a program of the following form: (3) Radiator Temperature (affected by mission envi- ronment). (1) Determine the total program of information neces- sary to derive the required parametric relation- (4) Thermal Tolerance Level (system-oriented con- ships. straint).

(2) Collect and assess the total information available. (5) Fuel Capsule Temperature Limit (safety-oriented constraint). (3) Determine the information remaining to be ob- tained. The selected thermoelectric hot junction temperature of item number A full program of parametric relationships is necessary 1 must be consistent with the numbers not only for RTG sizing, but also for indicating the selected for the other four items. If, for example, the hot major design problem areas, such as radiation and ther- junction temperature exceeds the allowable fuel capsule temperature limit, the feedback process requires that mal shielding, impact attenuating features, structural supports, etc. An understanding of those interface con- new numbers be placed into the matrix. Either the mod- ule design temperature should be lowered, or a different straints whose limitations result in the greatest design capsule technology, reflected by a higher temperature size and weight penalties will make possible their isola- capability, should be considered. The former may, in tion for purposes of special attention and development. turn, require the selection of a new thermoelectric mate- rial, and hence, new parametric numbers for the matrix. For practical utilization of the matrix of Fig. 1, num- The latter may require different fuels, and hence, new bers should be assigned to the blocks of the mission parametric numbers for the safety-oriented constraints. power source requirements. A thermoelectric material is then selected, and a design envelope of element sizing, number of elements, packing density, and operating tem- perature calculated. Numbers are next assigned to the VIII. Hybrid Systems generator oriented constraint blocks, and a check made The integration and feedback technique required for to see that the numbers assigned to the thermoelectric RTG sizing and analysis becomes further complicated module parameters do not exceed these figures. when considering shifts from a pure RTG power source to a hybrid system containing an RTG for continuous Assignment of numbers should then be made to all of power and a battery package for short duration peak the blocks in the matrix. Wherever possible, the numbers power. The benefits of such a system are indicated in should reflect measured hardware data; where no num- Fig. 2, where certain basic assumptions are made for a bers are available, best estimates should be used for an battery recharging rate and for an RTG power-to-weight initial rough cut. One of the functions of the matrix is to ratio. Figure 2, however, includes only electrical power point out those areas where data is either nonexistent, or characteristics, and does not factor in such important at best, rather sparse. considerations as power conditioning, trickle charging

JPL TECHNICAL MEMORANDUM 33-321 5 circuitry, and structural packaging necessary to put to- The module sizing process, although for exact analysis gether a hybrid unit. The mission power profile require- requiring a knowledge of all the assembly artifacts such ments should serve as a guideline in determining the as insulations, expansion compensators, etc., where trade-offs between pure RTG and hybrid systems. needed, can be roughly determined from the following:

Packing density as a function of element geometry Figure 2 illustrates the potential weight savings for a (4) and size. series of hybrid power systems sized from 50 w to 500 w peak power. The curves indicate that the shorter the (5) Thermoelectric array size and weight as a function duration of peak power requirement, the greater the po- of element packing density. tential weight savings for the system. This results from longer battery charging periods, hence the use of lower (6) Thermoelectric array size and weight as a function nominal power RTGs. of the number of couples.

Two other significant relationships are indicated by (7) Thermoelectric couple size and weight as a func- Fig. 2. The first is that the lower the peak power require- tion of element sizing. ment, the less the savings in weight by using a hybrid system; and second, the greater the ratio of nominal Since power output of thermoelectric couples is re- power required to peak power, the less the savings in lated to the electrical resistivity, and electrical resistivity weight by using a hybrid system. Thus, there exist power to temperature and geometry, the following relationships requirement limitations below which the power systems should be considered for varying temperature differen- engineer must analyze very carefully the tradeoff be- tials for a specific material: tween limited weight savings with a hybrid system and the reliability uncertainties involved with such a system. (8) Power output as a function of element geometry. (9) Output voltage as a function of the number of Figure 2 indicates that for high power requirements couples. of over 200 w, hybrid systems can produce as much as (10) Module specific power (w/lb) as a function of 10 w/lb of power system if the power profile requires a element packing density. peak power at least 10 times greater than the nominal power, and for only 10%)or less of the cycle duration. The following relationships indicate the extent to which mission requirements affect thermoelectric device IX. Parametric Relationships performance. The inputs provide the basis for sizing iterations: Basic parametric relationships are easily established in the area of thermoelectric module design. For any given (11) The effect of long life on voltage degradation thermoelectric material (a variable itself), the following (diffusion effects). relationships can be constructed: (12) The effect of long life on current degradation (1) Thermoelectric characteristics (Seebeck coeffi- (time versus resistance increase). cient, electrical resistivity and thermal conductiv- (13) The effect of thermal cycling on current degrada- ity) as a function of temperature. These curves are tion (thermal cycling versus resistance increase). available from the producers of thermoelectric material, and can be submitted to computer pro- (14) Reliability as a function of the number of couples gramming to translate the characteristics from and their series-parallel arraying. functions of temperature levels to functions of (15) Reliability as a function of element sizing. temperature ranges (or differentials). (16) Reliability as a function of stress and temperature (2) Open circuit voltage as a function of temperature levels. range (differential). (17) Reliability as a function of fabrication techniques. (3) Material efficiency as a function of temperature (18) Junction temperature changes as a function of differential. load variations.

6 JPL TECHNICAL MEMORANDUM 33-327 The thermoelectric device, unfortunately, is not with- (28) Thickness and amount of shielding material re- out limitations; and the inputs required for module siz- quired as a function of attenuating effect, for ing must be compared with the generator’s capabilities given fuel inventories of each isotope and fuel and limits being fed back into the integration analysis form. when design discrepancies occur. Parametric relation- ships 11, 12, and 13 are strong examples of important (29) Separation distance required between isotope feedback areas. Other limiting relationships are: fuel and target as a function of attenuating effect. (30) Structural weight as a function of separation Effect of impact loads on thermoelectric power distance. output.

Impact strength as a function of element sizing The dependence of the above relationships on reliable and geometry. isotope data inputs demands consistent integration of the system-oriented and safety oriented constraints. The ra- Impact strength as a function of element packing diation tolerance level of spacecraft science and elec- density. tronics may be the limiting factor in isotope selection. The availability affects not only the isotope and fuel Module size trade-off as a function of impact form, but also the half-life, power profile, and equipment resistance gain, for each specific attenuating necessary for power flattening, if a short half-life isotope feature. is selected.

Thermoelectric module weight trade-off as a Two complications exist in attempting to parametrically function of impact attenuating gain, for each spe- analyze the radiation relationships and problems. The cific attenuating feature. first is that selective or shadow shielding may be pos- sible. In place of a total system radiation shielding At present, parametric information on impact strength envelope, the more sensitive electronics and science and impact attenuating features such as resilient packing components may be individually shielded, the weight material, structural insulations, and couple supporting penalty for small component shielding being far less. In devices is almost non-existent. A testing program to ob- addition, no clear cut definition of radiation damage has tain the information would include a series of impact yet been established. The possibility exists that some tests starting at a nominal level of 100 g and increasing components, such as semiconductors, although “dam- by 100 g increments, each test including thermoelectric aged’’, by current usage terminology, may still function couples of three or more geometries, mounted in all four at levels sufficient for acceptable performance for the directional orientations. mission duration.

The result of system-oriented constraints is manifested The size and weight penalties paid for the use of in two ways: (1) directly increasing RTG size and weight thermal insulations may be directly traceable to the in such areas as thermal shielding and radiation shield- thermoelectric module design and the fuel loading tech- ing, and (2) indirectly influencing RTG size and weight nique. Some parametric considerations are: by affecting the thermoelectric module temperature (31) Surface area of inter-couple insulation as a func- range. Parametric relationships to be determined include: tion of couple packing density. (24) Types and amounts of science and electronics on- (32) Surface area of flat plate-RTG insulation (which board as a function of mission requirement. is approximately equal to the area of the housing) as a function of power level. (25) Radiation tolerance levels of science and elec- tronics. (33) Thickness of insulation as a function of tempera- ture drop (for heat fluxes from 0.1 to 10 w/in.2), (26) Type and amount of radiation as a function of for each insulation. isotope quantity, for each isotope and fuel form. The final area of parametric consideration, thermal (27) Quantity of isotope fuel required (fuel inventory) interfaces, touches all facets of RTG integration analysis. as a function of mission power requirement. Variations in thermal ambient conditions due to mission

JPL TECHNICAL MEMORANDUM 33-321 7 requirements such as sterilization, launch, entry, and where: planetary or lunar day-night cycling result in changes in temperature throughout the RTG. External barriers to 7 = overall RTG efficiency, "/o the rejection of RTG heat, such as sterilization shrouds and high entry velocity frictional heating, result initially and in increasing the radiator temperature. The limiting allowance in thermal ambient variation occurs when an Q = AeuTt + P, excessive radiator temperature rise is reflected by a hot junction temperature rise to the point where thermoelec- tric module failure occurs. Combining terms,

Compensation for temperature variations requires not Q - p, = -p, - Po = P" (f - 1) only an analysis of radiator sizing to meet the heat trans- B fer requirements, but also a reevaluation of the thermo- electric module design based on the potential fluctuations However, in heat flux. Q - Po = AEUT~ HTGs may be reduced to three basic elements: the heat source, the thermoelectric module, and the heat and exchanger. Assuming an ideal case, all of the isotope- generated heat passes into the thermoelectric array and, PO rl __ - __ (&UT;) with the exception of that portion converted to elec- A 1-r] tricity, is rejected at the module cold junction, T,. For design purposes, the hot junction, T,,, is considered a cons tan t . The above equation indicates that the power output of an KTG per unit radiating surface area is proportional to the fourth power of the radiator surface temperature, If the amount Of heat from the is Q, and the and roughly proportional to the RTG efficiency if it is thermoelectric power output of the module is P,,, small (approximately lox or less). the thermal energy left for rejection by the radiator is (I - P,).By the Stefan-Roltzmann law this heat rejection, Unfortunately, two opposing effects are dependent per unit of radiating surface is: upon the radiator temperature. Increasing the radiator temperature (T,) increases the heat rejection rate, and to a point, the power output per unit radiating surface area. On the other hand, increasing T, while TJtremains constant decreases the voltage, power, and efficiency of where: the thermoelectric module. This is not to imply that a change in a radiator temperature will not be reflected by a change in hot junction temperature for an operat- F = thermal emissivity of radiator, dimensionless ing device; it will, since the nuclear heat source is a a u = Stefan-Boltzmann constant, 5.67 X lo-" w/cm2 OK1 constant heat flux rather than constant temperature device. This is mentioned only to point out that once

% T = radiating surface temperature, 'K a hot junction design temperature has been selected, the designer is limited in his selection of a cold junction A = radiating surface area, cm? temperature to a trade-off optimization between module performance and radiator size and weight. and the power output of the RTG per unit radiating surface area is shown by The relationship between efficiency and radiator tem- perature is virtually inversely linear. By combining the --P" - Q various effects, the relationship between power output A '2 per unit radiating surface area and efficiency can be

8 JPL TECHNICAL MEMORANDUM 33-321 established. Other parametric curves involved with with all other components. Confidence in RTG sizing radiator sizing include: will be upheld only when data is available to sub- stantiate these numbers. Radiator surface area as a function of radiator temperature, per given power output. One word of caution is in order here, and that is that not all factors in RTG sizing lend themselves to para- Radiator weight as a function of surface area, per metric translations. Many of the AEC safety require- given material and thickness. ments, such as controlled intact reentry, may result in hardware additions of retro-rocket blow-off mechanisms, Radiator size and weight as a function of surface recovery aids, ablation shields, or stabilization fins, area, per given radiator geometry. whose weights, at best, can only be estimated. The same is true with RTG modularizing, locating, and orienting, where spin or three-axis stabilization requirements are X. Conclusion not readily determinable on a parametric basis. The end result of the total system analysis should be the assignment of size and weight numbers to all of the Appended to this document is a glossary of common components that make up the RTG. This should be done thermoelectric definitions and symbols to be used as a not on the basis of absolute component capability, but reference in examining the available thermoelectric liter- on the basis of relative component capability in concert ature.

JPL TECHNICAL MEMORANDUM 33-32 1 9 Appendix

Thermoelectric Definitions

1. Thermoelectric Material Effects and Parameters G. JouleHeat A. Seebeck Effect The thermal energy resulting from the Joule effect. The generation of an EMF resulting from a difference in temperature between the junctions of a circuit com- H. Figure-of-Merit of a Thermoelectric Material posed of two homogeneous conductors of dissimilar com- An algebraic factor indicating the proclivity of a ther- position. moelectric material for converting heat into electricity; algebraically, the quotient of the square of the absolute B. Seebeck EMF Seebeck coefficient and the product of the electrical re- (Also called Seebeck voltage, thermal voltage, open sistivity and the thermal conductivity, in dimensionless circuit voltage.) The EMF resulting from the Seebeck units/”C or OK. effect; the product of the Seebeck coefficient and the temperature difference between the circuit junctions, in 1. Figure-of-Merit of a Thermoelectric Couple volts. An algebraic factor indicating the proclivity of a ther- moelectric couple, composed of an N-type and P-type leg C. Seebeck Coefficient of specific materials, for converting thermal energy into An algebraic factor indicating the proclivity of a electrical energy. Algebraically, equal to given material to produce an EMF at any given tem- perature, in ,UV/~C. 1. Absolute. The integral, from absolute zero to the given temperature, of the quotient of the Thompson J. Figure-of-Merit of a Thermoelectric Couple, ‘Ideal coefficient of the material divided by the absolute EssentiaIly the same as above, but average values temperature. are used in the equation, the averages being obtained 2. Relative. The coefficient of a couple composed of a by integrating the parameters over the operating tem- given material and a specified standard material, such perature range of the couple. as platinum or copper. 3. Of a couple. The algebraic difference of either the relative or absolute Seebeck coefficients of the two II. Performance Criteria conductors. A. Power Output

D. Peltier Effect The rate of electrical energy produced by a thermo- electric device, w. In a circuit composed of two dissimilar conductors carrying current, the evolution of heat at one junction of the circuit and absorption of heat at the other junction. B. Efficiency A performance factor equal to some specific output E. Peltier Heat quantity divided by a specific input quantity, in percent. The thermal energy absorbed or evolved as a result of 1. Overall generator. The ratio of electrical power out- changes in the current in the circuit (and, hence, in the put to total consumed thermal input. Peltier effect). 2. Module. The ratio of electrical power output of a F. Joule Effect therma$ectrical set to its utilized thermaI power input. The creation of thermal energy in a current-carrying 3. Carnot. The ideal thermal efficiency of a heat en- conductor resulting from the Joule effect. gine, equal to ( Th - T,)/Th.

10 JPL TECHNICAL MEMORANDUM 33-321 C. Specific Power 111. Thermoelectric Hardware - Heat Sources The ratio of electrical power output of a thermoelec- A. Heat Source tric device to its weight, w/lb. Any source of thermal energy to be converted into D. Voltage electrical energy, such as solar energy, fossil fuels, nu- clear reactors, etc. 1. Open circuit. The generated Seebeck voltage mea- sured across the unloaded, open terminals of a thermo- electric device, v. B. Heat Transfer Mode 2. Load voltage. The voltage drop measured across The means by which energy from the heat source is the terminals of a device with a current flowing. transmitted to the thermoelectric set.

3. Matched load voltage. The voltage drop measured 1. Thermal conduction. The transmission of heat from across the terminals of a thermoelectric device when the one part of a body to another part of the same body, or resistance of the current-causing load is exactly equal to from one body to another in physical contact with it, the internal resistance of the device. without displacement of particles of the body, and at a rate proportional to the temperature gradient of the body. E. Current 1. Load. The dc current flowing through a thermo- 2. Thermal convection. The transfer of heat from one electric device when the generated voltage is impressed region to another within a gas or liquid by the mixing of across a given resistive load. one portion of the fluid with another.

2. Short circuit. The dc current flowing through a 3. Thermal radiation. The transfer of heat by the thermoelectric device when its output terminals are short emission of radiant energy from one body and the ab- circuited. sorption of that energy by another body of lower tem- perature. F. Internal Resistance The total resistance of the thermoelectric circuit of a C. Heat Input thermoelectric device, measured between the output ter- minals, and including the thermoelectrical materials, the The total thermal power delivered by a given quantity electrical connectors, and the thermoelectric couple of a heat source. contacts.

G. Contact Resistance D. Useful Heat Input The resistance measured across the interface where The total thermal power actually absorbed by a ther- the thermoelectric material is contacted to a shoe. moelectric set.

H. Degradation E. Temperature The ratio of the change in a measured electrical pa- 1. Hot junction. The temperature, in degrees centi- rameter at any given time to the initially measured elec- grade, at the contact interface between the thermoelec- trical parameter, measured consistently at operating tric material and the hot shoe. temperature or at ambient temperature, in percent. The change will be an increase for internal resistance, and a 2. Hot side. The temperature, in degrees centigrade, decrease for current or voltage. measured in the region of the hot junction, such as at the hot shoe or in the thermoelectric element, but not 1. Failure actually at the hot junction. The occurrence of a complete open within the thermo- electric circuit, a complete short between the terminals 3. Hot wall. The temperature, in degrees centigrade, within the circuit, or degradation to some preestablished measured at the surface of the hot wall facing the ther- level. moelectric set.

JPL TECHNICAL MEMORANDUM 33-321 11 F. Hot Wall E. P-Leg The device component sandwiched in between the A thermoelectric leg made from thermoelectric mate- heat source and the thermoelectric set for the specific rial with P dopant added. purpose of receiving heat from the source and transfer- ring it, either by conduction or radiation, to the thermo- F. N-leg electric set. A thermoelectric leg made from thermoelectric mate- rial with N dopant added. G. Hot Frame

The structural member, surrounding or adjacent to the G. Thermoelectric Set heat source, used to support the thermoelectric set on Any general array of interconnected thermoelectric the hot side. legs, couples, or modules.

H. Hot Side H. Thermoelectric Module Any section of a thermoelectric device which includes A structurally sound and fully assembled array of two any or all components from, and including, the hot junc- or thermoelectric couples, tion to the center or centerline of the heat source.

1. Thermoelectric Generator 1. Hot Junction A complete thermoelectric power package, with oper- The contact interface plane between the thermoelec- ational capability, including the heat source, the themo- tric material and the hot shoe of a thermoelectric device. electric array, and the heat exchanger,

J. Thermoelectric Device IV. Thermoelectric Hardware - Thermoelectric Assemblies Any thermoelectric hardware, from a couple to a generator. A. HotShoe (Also called contact, cap, electrode, connector.) The K. Insulation electrically conducting member, either metallurgically or 1. Electrical. The dielectric material used to prevent mechanically in contact with the hot side of a thermo- individual thermoelectric couples in a circuit from elec- electric element, to which an electrical connection may trically shorting with each other when the thermoelectric be made with an adjacent member to form a portion or set is in contact with an electrically conducting common all of a thermoelectric circuit. hot or cold wall.

B. Thermoelectric Element 2. Thermal. Material of low thermal conductivity used to prevent thermal energy from the heat source from by- (Also called pellet.) An uncontacted piece of processed passing the thermoelectric set. thermoelectric material fabricated into a definite geo- metrical configuration. 1. Cascading

C. Thermoelectric leg Aligning two or more thermoelectric devices thermally in series, with the cold side of the first stage device, the (Also called member.) A thermoelectric element which hot side of the second stage device, etc. has been contacted on both the hot and cold ends; the main component of a thermoelectric couple. M. Segmenting The fabrication of a thermoelectric leg with D. Thermoelectric Couple two or more distinct and finite regions between the hot and An electrically connected pair of thermoelectric legs, cold junctions having either different compositions or one P doped and one N doped. dopant levels.

12 JPl TECHNICAL MEMORANDUM 33-327 V. Thermoelectric Hardware - Heat Rejection J. Temperature A. Heat Exchanger 1. Root. The temperature of a radiating fin at its base, "C. The heat transfer mechanism for removing heat from a thermoelectric generator, including radiator fins, con- 2. Radiator temperature. The average temperature, vection fins, and liquid metal loops. measured in degrees Centigrade, of the radiating fin.

B. Cold Wall 3. Cold side. The temperature, "C,in the region of the cold junction, such as at the cold shoe or in the ther- A device sandwiched between the moelectric leg, but not actually at the cold junction. side of the thermoelectric set and the heat exchanger. 4. Cold junction. The temperature, "C, at the contact 6. Cold Strap interface between the thermoelectric material and the (Also called cold connector, bridge.) A metallic ther- hot shoe. moelectric circuit member electrically connecting adja- cent N or P thermoelectric legs. VI. Nuclear Applications D. Cold Shoe A. RTG (Also called contact, cap, electrode, connector.) The Radioisotope Thermoelectric Generator. electrically conducting member, either metallurgically or mechanically in contact with the cold side of a thermo- 8. SNAP electric element, to which an electrical connection may be made with an adjacent member to form a thermoelec- Systems for Nuclear Auxiliary Power. tric circuit. C. Radioisotope E. Cold Frame Radioactive isotope of an element. The structural member, internal to or adjacent to the heat exchanger, used to support the thermoelectric set D. Half-life on the cold side. The time period required for disintegration of one-half the atoms of a radioisotope. F. Cold Side

Any section of a thermoelectric device which includes E. Alpha Emitter any or all components from, and including, the cold junction to the extremities of the heat exchanger. A radioisotope whose decay results in a preponderance of alpha particle radiation, which, because of its large mass and ionized state, needs relatively minor shielding. G. Cold Junction

The contact interface plane between the thermoelec- F. Beta Emitter tric material and the cold shoe of a thermoelectric device. A radioisotope whose decay results in the simultaneous emission of an electron and a neutrino, with the sum of H. Fin Efficiency the energies of the two emissions being equal to the total A decimal number indication of the potential of a heat beta decay energy for that particular transition. exchanger fin to maintain a uniform temperature along its length. G. Gamma Emitter A radioisotope whose decay leaves the product nu- 1. Effective Radiating Area cleus in an unstable energy level, subsequently decaying The area of a prismatic surface just enclosing a finned to the ground state with the emission of one or more generator with its fin length reduced by the fin efficiency. photons.

JPL TECHNICAL MEMORANDUM 33-32 7 13 H. Tolerable Radiation level R. Specific Power The level of radiation which will be harmful, if ex- The amount of thermal power produced per unit mass ceeded, to personnel, electronics, and scientific experi- of radioisotope, w/g. mentation associated with the nuclear-fueled generator.

1. Shielding VII. Mission and Application Considerations The requirement for reduction of raw nuclear radia- A. Mission tion to acceptable levels through the use of radiation A preplanned and fully integrated space program or attenuation materials placed between the nuclear source project in which performance of the power is only and vulnerable targets. a portion of the total objective.

J. Shadow Shielding B. Application (as opposed to a:mission) The use of partial radiation shielding to protect only A space project whose primary objective is perfor- those areas of extreme radiation sensitivity. mance of the power source; or any terrestrial program or project using an auxiliary power source, but whose ob- K. Fuel Capsule jective is the production of electrical power for specific The sheath or enclosure within which the radioisotopic reasons. fuel is contained. C. Beginning of life 1. Fuel Form The initial time, during a mission or application, when The chemical compound and/or the physical condition the thermoelectric generator first reaches design tem- of a radioisotopic fuel. perature or becomes operational.

M. Intact Reentry D. End of life The nuclear safety philosophy which requires com- The point in time when the mission or application for plete containment or retention of nuclear fuels following which the thermoelectric generator was intended is reentry of a generator into the earth's atmosphere and ended, even though the generator itself may still be onto the earth's surface. functioning.

N. Burn-up Reentry E. Power Profile The nuclear safety Philosophy Which requires the con- The pattern of power level (w) and power duration sumption of nuclear fuels by aerodynamic heating and (hours) requirements for a given heperiod. subsequent product dispersion due to atmospheric re- entry of a generator. F. Reliability 0. Fuel Inventory The statistical expectation for a complete thermoelec- The quantity of nuclear fuel in a generator, in curies. tric power system to meet performance specifications for a full mission or application duration. P. Power Flattening G. Mission Environment The technique of maintaining a constant electrical power output from an RTG fueled by a nuclear heat 1. lhd iWaCt. The requirement for an RTG to Sur- source with a relatively short half-life. vive an impact of over 1000 g at greater than 150 ft/sec velocity. Q. Power Density 2. Orientation. The direction of the radiator fins of an The amount of thermal power produced per unit vol- RTG relative to a clear view of space and any obstruc- ume of radioisotope, w/cm3. tions to that view.

14 JPL TECHNICAL MEMORANDUM 33-321 3. Atmosphere. The pressure, temperature, and fluid preplanned period of hours until all contaminating or- velocity external to the RTG at any given time after ganisms have been thermally destroyed. beginning of life. 1. interface Constraints In a mission context, the design limitations with which H. Heat Sterilization the RTG must live with to be fully compatible electrically, The requirement that the RTG be enclosed in a her- thermally, volumetrically, weight-wise, and radiation- metically sealed shroud in which the ambient tempera- wise with the payload, the spacecraft, and the ultimate ture will be held at some discrete temperature for a missionobjectives.

Nomenclature

a relative Seebeck coefficient for a couple composed c cold junction (T,) of legs P and N ca carnot (Nca) x relative Peltier coefficient for a couple composed cs cold side ( Tes) of legs P and N co couple (Veo,Reo, Ne,, wee) P resistivity, a-cm R resistance, a e electrical (Kw,, E,) k thermal conductivity, w/cm C f final (tr, Rr, N,) K thermal conductance, w/ C in internal (Rim,Vin) V voltage, v i initial (Ti, RJ,Ni) I current, amp i joule (QJ

T temperature, O C 1 load (Vz, Iz, Rz) AT temperature difference between the hot and cold hs hot side (Ths) junctions h hot junction (Th) P electrical power, w n N-leg (V,, a,, W,, VA,) quantity of heat, w Q oc open circuit (Voc) 7 efficiency, % 0 output (Po) z figure-of-meritof a thermoelectric material P P-leg (VP, RP, WP,VAP, ap, kP> P’ specific power, w/lb sc short circuit (Ise) E energy, w/hr t thermal (Kwt,Et) t time (hr, day, etc.) T relative (a,) W weight, (g, lb) rn module (vm) 1/A length over area ratio, cm-1 ml matched load (GZ,Vmz> Subscripts x loss (Qo&) a available (Qa,E,) I: total, overall (Q ,P , 7 )

JPl TECHNICAL MEMORANDUM 33-32 1 15