DIRECT CONVERSION FOR SPACE SOLAR POWER

An Undergraduate Thesis
Presented to
The Academic Faculty

by

Nicholas Boechler

In Partial Fulfillment
of the Requirements for the Degree
Bachelor of Science in Aerospace Engineering, Research Option in the
School of Aerospace Engineering, College of Engineering

Georgia Institute of Technology

May 2007
DIRECT CONVERSION FOR SPACE SOLAR POWER

Approved by:

Dr. Narayanan Komerath, Advisor

School of Aerospace Engineering
Georgia Institute of Technology

Dr. Erian Armanios

School of Aerospace Engineering
Georgia Institute of Technology

Date Approved: April 27, 2007

 ACKNOWLEDGEMENTS

I wish to thank Dr. Komerath, my research advisor. It has been a true honor, to

work with him in his research endeavors. In addition to the great opportunity he has given

me, he has always listened to my ideas, and given me his confidence, respect,

encouragement, and support. I am forever indebted. I would like to thank my mother and

father for everything. I would like thank Dr. John Olds for the all the opportunities, help,

and advice he has given me over the years. I would like to thank Dr. Armanios for

reviewing my thesis, along with the recommendations and advice. Thanks to Dr. Donnell

for thesis writing instruction and advice. Thank you to the NASA Institute for Advanced

Concepts, SAIC, and the Georgia Tech Undergraduate Research Opportunities Program

for funding my work. Thanks to the Georgia Institute of Technology. Thanks to Dr.

Joannopoulos for discussing photonic crystals. And last but not least, I would like to

thank my friends, who have tolerated my constant barrage of wild ideas.

iii
 TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS iii

LIST OF TABLES vi

LIST OF FIGURES vii

LIST OF SYMBOLS AND ABBREVIATIONS viii

SUMMARY ix

CHAPTER

1 Introduction 1

2 Literature Review 3

2.1 Historical Space Solar Power Systems 3

2.2 Roadblocks to Space Solar Power 4

2.2.1 Current Photovoltaic Solar Cell Technology 4

2.2.2 High Launch Costs 6

2.2.3 Atmospheric Transmission 7

2.3 Technologies for Direct Conversion 8

2.3.1 Signal Processing Solutions 8

2.3.2 Rapidly Ionizing Plasma 11

2.3.3 Nanofabricated Antenna 11

2.3.4 Optical Resonator 12

2.3.5 Shocked Photonic Crystals 13

2.3.6 Solar Pumped Lasers and Masers 14

2.3.7 Optical Rectenna 17

2.4 Conclusion of Literature Review 19

iv
 3 System Concepts 20

3.1 General System Components 20

3.2 Shocked Photonic Crystal System 22

3.2.1 Concept Description 22

3.2.2 Estimate 24

3.2.3 Implementation Analysis 25

3.3 Solar Pumped Maser System 26

3.3.1 Concept Description 26

3.3.2 Estimate 28

3.3.3 Implementation Analysis 30

3.4 Optical Antenna System 32

3.4.1 Concept Description 32

3.4.2 Estimate 35

3.4.3 Implementation Analysis 35

3.5 Conclusion of Systems Concepts 36

4 Selected Applications 38

4.1 Space Solar Power Grid 38

4.2 Electric Propulsion 40

4.3 Conclusion of Selected Applications 42

5 Conclusion 43

REFERENCES 45

v
 LIST OF TABLES

Page

Table 1: Space Rated Si and GaAs Solar Array Performance Survey8-10 6

Table 2: Launch Cost Survey12-14 7

Table 3: Sample astronomical maser transitions and frequency output23 27

Table 4: Summary of direct conversion system concepts specific power estimates and
conclusions 37

Table 5: Cost comparison of direct conversion concept implementation on a 125 MW

solar power satellite 39

Table 6: Direct Conversion for SSP summary table 44

vi
 LIST OF FIGURES

Page

Figure 1: Atmospheric transmission as a function of frequency15 8

Figure 2: Simulation of Frequency Shift in a Shocked Photonic Crystal by Dr.

Joannopoulos20 13

Figure 3: Model of 3D Photonic Crystal by Dr. Joannopoulos19 14

Figure 4: Diagram of W.C. Browns Thin Film Etched Circuit Rectenna27 17

Figure 5: Shocked Photonic Crystal System Concept Diagram 23

Figure 6: Solar pumped maser system concept diagram 27

Figure 7: Cylindrical aspect ratio relation to maser saturatio23 30

Figure 8: Optical antenna and rectenna concept diagrams 34

Figure 9: Simulation of low frequency wave approximation by 2x max frequency

sampling 34

vii
 LIST OF SYMBOLS AND ABBREVIATIONS

SSP Space Solar Power

DC Direct Current

AC Alternating Current

EM Electromagnetic

RF Radio Frequency

SAW Surface Acoustic Wave

Laser Light Amplification by Stimulated Emission of Radiation

Maser Microwave Amplification by Stimulated Emission of Radiation

viii
 SUMMARY

Space Solar Power (SSP) is a powerful yet nearly untapped resource with

revolutionary potential. SSP systems currently have several roadblocks to their

implementation. With the technology in use today, converting solar power to useable

energy is inefficient, the required converters have a large mass per unit power, and

launching those converters is expensive. More fundamentally, in all current SSP systems,

energy is generated in the form of a direct current before being converted again into

whatever form is necessary. In addition to the large mass per unit energy of this

conversion equipment, such conversion involves significant efficiency losses, further

resulting in the prohibitive cost of launching these converters into space. If techniques

could be discovered for converting broadband sunlight directly to a useable narrowband

application dependent frequency, many fundamental breakthroughs in aerospace

endeavors can be achieved.

This project studied a large number of options that might lead to direct

conversion. Those technology options were analyzed according to which would warrant

further exploration from the point of view of aerospace systems applications and possible

power per unit mass. Based on these technologies, several advanced concepts were

considered. It is also important to make an estimate of the possible power per unit mass

that could be achieved with each concept, so that architecture developers can proceed

with the development of applications enabled by direct conversion technology.

Accordingly, estimates of the possible power per unit mass of potential direct conversion

ix
systems were made, and future applications that would benefit from those direct

conversion systems were identified.

Three possible concepts were developed. These concepts include: a shocked

photonic crystal system; a solar pumped maser based on naturally occurring astronomical

masers; and an optical antenna array with central signal processing. The optical antenna

array and the solar pumped maser were estimated to have a specific power approximately

15.0 and 10.8 times greater, respectively, than conventional photovoltaic systems.

Several applications were identified that would benefit from direct conversion

systems. The most obvious would be an orbital SSP grid. In addition to the rise in

specific power, and therefore resulting cost savings, direct conversion would enable the

output of higher frequency transmissions with less beam spreading and subsequently

smaller ground based infrastructure. Electric propulsion systems would also benefit

directly by drastically reducing on board mass both from higher specific power and direct

conversion to the necessary ionization frequency. Additionally, the inclusion of solar

sails opens possibilities for possible hybrid designs incorporating integrated propulsion

and energy collection.

As with any sufficiently advanced concept, it is impossible to be certain where the

problems, roadblocks, and successes may lie in the future. This said, this project has

shown that several direct conversion concepts warrant further exploration and study

based on their revolutionary potential. They are a true shift from the paradigm of

conventional SSP.

x
 CHAPTER 1

INTRODUCTION

Space Solar Power (SSP) is a powerful yet nearly untapped resource with

revolutionary potential to address the energy needs and environmental concerns of

humanity. SSP systems currently have several roadblocks to their implementation. With

the technology in use today, converting solar power to useable energy is inefficient, the

required converters have a large mass per unit power, and launching those converters into

high orbits is prohibitively expensive. More fundamentally, in all current SSP systems,

captured energy is converted to direct current (DC) before being converted again into

whatever form is necessary. If techniques could be discovered for converting broadband

sunlight directly to a useable narrowband application dependent frequency, many

fundamental breakthroughs in aerospace endeavors can be achieved.

This study will explore, propose, and analyze possible concepts that would

provide great benefits to SSP systems through the use of direct conversion. The process

and goals include:

I. Identify Applicable Technologies: Explore several options that might lead to direct

conversion. No single solution might be the answer, so several types of conversion

technology must be explored.

II. Develop System Concepts: Following the analysis of applicable technologies, a few

space solar power direct conversion concepts will be developed.

1
III. Analyze for Aerospace Applications: In preliminary review of this field, it is

evident that little attention has been paid to high power/low mass systems. Even after

finding a technology, and proposing a related concept that would warrant further

exploration, it must be looked at from the point of view of aerospace systems applications

and possible power per unit mass. The key is to develop estimates, however rudimentary,

of the possible power per unit mass achievable with each concept, as this is the most

significant metric for aerospace systems intended for launch into orbit.

IV. Estimate Specific Power: Provide a justifiable estimate of power per unit mass of

future direct conversion systems. This will further enable the development of concepts

involving direct conversion simultaneously to the development of direct conversion

technology itself.

V. Future Applications: Identify possible future applications that would benefit from

direct conversion technology.

2
 CHAPTER 2

LITERATURE REVIEW

There exists significant background information as to the general concept,

advantages, and problems involved with historical SSP systems. This background

information is necessary to establish the rationale for developing direct conversion

systems. There are also many natural and technological analogs to the direct conversion

process. These technologies will be adapted for use in, and the development of, SSP

direct conversion concepts.

2.1 Historical Space Solar Power Systems

For decades, many space solar power (SSP) concepts have been proposed that

show great potential for revolutionizing energy production on Earth.1-6 SSP in general has

several advantages. As fossil fuel supply dwindles, and concerns about pollution and

global warming heighten, the clean, plentiful energy from the sun becomes increasingly

attractive. Satellites in high orbits receive intense solar energy around the clock,

regardless of weather or seasonal variations. Additionally, the essentially infinite area of

space is available for solar arrays, in contrast to the relatively limited area on Earth.

Finally, in some SSP concepts, the collected power can be transmitted to multiple sites

around the Earth.4-6

3
 2.2 Roadblocks to Space Solar Power

Despite the great potential, several practical and technological roadblocks have

prohibited the implementation of SSP concepts. The two most critical problems, which

play into and further cause other areas of difficulty, are the nature of current photovoltaic

technology and high launch costs. Present-day photovoltaic technology is theoretically

limited in its achievable efficiency, and produces its useable energy in the form of DC

only, which is inefficient to transmit over cables and unsuitable for wireless beaming.

The production of DC also means that any SSP concept must also incorporate heavy

converters to transfer the collected power to Earth. This power transfer is commonly

accomplished through radio or microwave beaming, after conversion to alternating

current (AC).

The problems with conventional photovoltaic technology coupled with high

launch costs define the guidelines around which SSP direct conversion systems can be

designed. The system must convert broadband radiation directly to low frequency

narrowband radiation for transmission, and it must have a low mass per unit power. In

addition, this output frequency must be tailored for maximum atmospheric transmission.

A previous deterrent to higher frequency space transmissions is the availability of

atmospheric transmission windows. This is a problem in that generally lower frequency

implies higher beam spreading, which requires a huge ground antenna as a receiver.

2.2.1 Current Photovoltaic Solar Cell Technology

The current method for solar power generation is based upon essentially the same

technology as the original silicon photovoltaic cells. These cells work by using photons to

4
excite electrons in a semiconductor to produce an electric current. This method is

fundamentally limited for several reasons. Sunlight is broadband energy and only photons

with certain energy can dislodge an electron. To become part of the circuit the electron

must jump a band gap specific to the material. All the energy put into jumping this band

gap is converted to waste heat. Because of this, traditional photovoltaic cells already have

around a 55% theoretical efficiency loss.7

In addition, photovoltaic cells produce only direct current, and consideration must

be taken that many space applications require electromagnetic (EM) radiation or other

waveform currents. This re-conversion system would require extra mass and add an extra

factor for efficiency loss, compared to something that already has an outgoing alternating

current. When the initial power is generated from free space solar radiation, and the end

goal is another form of radiation for power transfer, it is logical that complexity and mass

could be saved by directly converting.

Table 1 specifies a range of solar cell performance values from literature. It is

important to note that this is simply a characteristic range of available solar arrays for

space use. There may be specific, newer technology, examples that may be outside this

range. However, this range serves well for estimation purposes. Additionally, new

photovoltaic concepts that may provide higher performance would theoretically still have

the same problem of generating power in direct current which would need to be

converted again for transmission.

5
 Table 1: Space Rated Si and GaAs Solar Array Performance Survey8-10

Beginning of Life Values Minimum Maximum

Specific Mass (kg/m2) 0.23 2.88
Efficiency 0.09 0.26
Specific Power (kW/kg) 0.12 1.01
Specific Power (kW/m2) 0.12 0.36

2.2.2 High Launch Costs

Probably the most critical roadblock to a presence in space is the high launch

costs. Because of this, when dealing with SSP systems the specific power of a system is a

make or break metric for the success of the system as a whole. This can be magnified

even further to the point that a SSP satellite with a given high specific power may not be

able to recuperate its launch costs from power sales over its lifetime, therefore making it

cost prohibitive. This launch cost problem is also further exacerbated by conventional re-

supply and stabilization problems.11 SSP satellites in orbit will continually require fuel

and navigation to correct perturbations to their orbits due to solar pressure, orbital decay,

and other destabilizing factors. Thus the launch cost for launching additional propellant

must also be taken into consideration, perhaps further decreasing the necessary specific

power. Some current prices for low earth orbit launches are shown in

Table 2.12-14 It is important to note that the cost of launching to geosynchronous

Earth orbit, which would be applicable to certain SSP concepts, is generally greater that

launching to LEO by a factor of two or more.

6
 Table 2: Launch Cost Survey12-14

Specific Cost
Payload LEO Cost ($K
(kg) ($M) Year FY07/kg)
Ariane 44L 22400 125 2000 3.1
Atlas V 12500 138 2004 11.1
Delta II 5089 60 1999 97.0
Delta IV Medium 8600 133 2004 16.7
Saturn V 118000 431 1967 3.1
Space Shuttle 24400 245 1988 83.6
Titan 21680 432 1999 6.5
Average 2007 31.6

2.2.3 Atmospheric Transmission

As evident in Figure 1, there are several frequency ranges where a large

percentage of transmission energy is lost in the atmosphere, mainly due to absorption by

water vapor. This is particularly important to this project in that the output frequency of

any viable system must be tailorable. It also defines the target frequencies for which the

direct conversion concepts will be designed.

7
Figure 1: Atmospheric transmission as a function of frequency15

2.3 Technologies for Direct Conversion

Several technologies have been identified as potential candidates for a direct

conversion system because of similar natural mechanisms that provide frequency shifting

and bandwidth narrowing. The direct frequency conversion from broadband radiation to

low frequency narrowband radiation could lower complexity and mass per unit power of

an SSP system. The specific natural and technological analogs that were analyzed in this

project include, signal processing solutions, interactions with rapidly ionizing plasma,

nanofabricated optical antennae and rectennae, optical resonators, shocked photonic

crystals, solar pumped lasers and masers.

2.3.1 Signal Processing Solutions

Signal processing solutions would initially seem to be the most obvious solution

to frequency conversion, but unfortunately none of these by themselves are very efficient

8
or much of an improvement from the existing photovoltaic plus microwave transmitter

system. Generally they are intended for low power applications, where issues such as low

conversion efficiency or heat generation do not affect their use. These signal processing

devices are common in all sorts of electronic devices used in every day life. A perfect

example is a travel AC frequency converter. This might convert 120 Hz power to 240 Hz

power. However the conversion efficiency is poor, and much energy is lost through heat.

All following devices are relatively common circuit configurations, of extrapolations

there upon, and therefore will not have specific citations. These examples were generated

through discussions with electrical engineering professors, and are important background

to rule out certain direct conversion possibilities.

Assuming that broadband sunlight could be coupled to an antenna, a configuration

of diodes in the form of a mixer powered by an oscillator could achieve a frequency shift.

The output would be the sum and difference of the input and oscillator frequencies, one

of which would need to be filtered out. The oscillator would also require its own power

source. Aside from the need for an undetermined quantity of power to be supplied to the

oscillator, this system would be inherently inefficient due to the incorporation of a filter,

and there is no obvious way to tailor this to a broadband input.

Other options include modulating the input signal with a radio frequency (RF)

signal, which is very inefficient. In addition there are many products on the market today

that provide a small range of frequency translation such as SAW filter circuits, but again

these have a very low translation range and efficiency.

Tube, Cyclotron, and Magnetron type devices also fall into this category. These devices

fundamentally work by producing an EM field by crossing a magnetic field with

9
electrons produced by a DC power supply from a cathode to an anode. These devices are

the current and most common way of producing microwave radiation. There are several

problems however with this setup in relation to SSP systems. These problems include

mechanical tolerances, scaling problems, quality decay over time, weight, impedance

mismatching, and breakdown fields. Most of these problems get amplified as higher

frequency output is attempted, further decreasing the efficiency. These higher frequency

issues would be especially pertinent to this application considering the success of many

of the applications depend on limited beam spreading during transmission which is

minimized at higher frequencies. In addition the whole concept of direct conversion

strives to eliminate just this type of DC to AC step.

An alternative method of signal processing may be useful to direct conversion

concepts. Assuming that high frequency broadband current could be analyzed in real

time, signal processing solutions could potentially reconstruct the high frequency part of

the signal into a higher intensity low frequency wave. This idea is similar to that of a

Fourier approximation, and should be theoretically possible as per the NyquistShannon

sampling theorem. The NyquistShannon sampling theorem specifies that a wave can be

characterized with sampling at twice the wave frequency. Although potentially difficult

because a system would be dealing with THz radiation, it should be theoretically

possible. Such a concept, although advanced, would work well with optical antenna array

systems. It should be noted here that the above need not be done with active digital signal

processing using computers. Optical Fourier Transform elements (a prism separating

various colors spatially is a simple example) are available, and can separate broadband

sunlight into discrete, narrow bands with high efficiency. Reconstructing these into a

10
narrow band in a frequency range far below that of visible light, is what poses the

challenge.

2.3.2 Rapidly Ionizing Plasma

Studies have shown that an interaction between EM radiation and rapidly ionizing

plasma can cause a simultaneous frequency up-shift and down-shift.16 This concept

seems particularly applicable. However, it would be difficult to develop a viable concept

considering that a large amount of energy would be required to continually re-ionize the

plasma.

2.3.3 Nanofabricated Antenna

EM radiation of lower frequencies (microwave and lower) has traditionally been

generated and then transmitted through antennae in the classical physics sense. EM

radiation of higher optical frequencies has been more difficult because of decreasing size

to the realm where quantum mechanics must be taken into consideration. Developing a

beam in the optical frequencies has been accomplished by lasers but is somewhat

inefficient due to the population inversion. If nanofabricated antennae were possible

perhaps radiation collection and emission could be enabled in the classical sense. This is

theoretically possible considering that ITN Energy Systems asserts that this can be done

with its optical rectenna with additional nanofabricated rectification components.17

Although ITN is only attempting it in the collection direction by absorbing the radiation,

it should be theoretically possible also to symmetrically re-emit it in a similar fashion

11
with tailored geometry for the emission of a focused beam. A nanofabricated antenna

such as this should be given focus as an initial step without rectification. Just coupling

broadband radiation to an antenna array would be an important breakthrough in itself.

This is important because it allows concepts to move away from traditional means of

microwave transmission such as cyclotrons and gyrotrons. Instead, assuming the signal

has already been converted to the necessary frequency and that nano-fabrication

techniques become practical in the near term, it can be transmitted in radiation form by a

simple and low mass mechanism.

2.3.4 Optical Resonator

Optical resonators take broadband light and convert it to an amplified narrow

band. Resonators have a clear benefit to applications needing EM radiation amplification,

but also might be used in direct frequency conversion. There is a Jet Propulsion Lab

proposal for a toroidal or disk-like optical resonator made of an optically non-linear

material for parametric frequency conversion.18 One of the main benefits of resonators in

general is the simplicity and low weight. A simple resonator can be made out of two

parallel plates of reflective material. The broadband frequency will approach a frequency

proportional to the length of the resonating cavity, along with the corresponding sub-

harmonics. However, resonators in general have the critical drawback that they work by

amplifying the resonant frequency and rejecting the non-resonant frequency. This non-

resonant frequency rejection thereby implies an energy and efficiency loss.

12
2.3.5 Shocked Photonic Crystals

In 2003 a group at MIT used a simulation to show that non-relativistic reversed

Doppler Shift with near 100% efficiency in light occurs when light is reflected from a

moving shock wave propagating through a photonic crystal, as shown in Figure 2.19

Figure 2: Simulation of Frequency Shift in a Shocked Photonic Crystal by Dr.

Joannopoulos19

The overall concept of photonic crystals and their applications is described in Dr.

Joannopoulos book Photonic Crystals: The Road from Theory to Practice.20 He

explains that photonic crystals, which can be fabricated on the nano scale and through the

creation of the correctly placed defects and tailored geometry, can create perfect resonant

cavities or waveguides depending on the design of the crystal structure. The efficiency of

such devices largely depends on the quality of fabrication, but simulations show near

100% efficiency in theory.19-21

13
Figure 3: Model of 3D Photonic Crystal by Dr. Joannopoulos19

The system potentially offers tunable pulse rate and carrier frequency based upon

artificial band gap size and can narrow the bandwidth of the incident radiation. This is

exactly the effect that is desired for direct conversion. After a significant amount of

iterations light could be dramatically shifted in frequency. Additionally, this system has

high theoretical efficiencies, and it was proposed that a similar system be used in micro-

electrical-mechanical devices.19-21 This technology will be utilized and further described

in the direct conversion system concepts.

2.3.6 Solar Pumped Lasers and Masers

Respectively, the acronyms stand for Light or Microwave Amplification by

Stimulated Emission of Radiation. Both function in essentially the same way. Energy

from the pumping mechanism, which is solar radiation in this case, causes excited states

in the molecules of the gain medium. Enough excitation causes a population inversion

and the emission of EM radiation as the particle falls to a lower energy state. This

14
transition can be excited by collisions with photons of the same energy as the difference

between the two states, so that each successful collision generates a photon at the same

energy and phase as the colliding photon. This amplification of the beam can be

increased by passing the beam multiple times through the cavity containing the lasing

medium, as occurs in a resonator. The net gain is the balance between the release of

energy at the right frequency during collisions with the stimulating photons, the losses

due to absorption of the photons at cavity walls or by molecules without emission, and

spontaneous emission at random phase. At high intensities, the spontaneous emission is

negligibly small compared to the absorption and the stimulated emission. While present-

day ion, and metal vapor lasers use electronic state transitions, masers use ro-vibrational

transitions. These transitions between two rotational energy levels associated with

different vibrational energy levels are also used in high-power gas lasers with mediums

such as carbon dioxide, water, and nitrogen. A critical argument can be made that

because of scale, maser transitions should be much easier to control and therefore

potentially more efficient. Also, empirical evidence on lasers shows that the efficiency

achievable with a laser is much higher at lower frequencies, such as infrared, than in the

visible or ultraviolet range. This observation suggests that masers operating in the 100

GHz regime, should in principle be capable of higher efficiencies than those

demonstrated with lasers. This concept is well suited for direct conversion in that a

broadband energy source can be re-emitted non-linearly in a narrowband form.

Natural analogs exist in the form of astronomical masers. Astronomers for some

time have been discovering bright spectral emissions created by naturally occurring

interstellar masers.22-23 This naturally occurring phenomenon is particularly suitable to

15
the application of direct conversion. Essentially a low density molecular gas is excited by

collisions from photons and molecules from a broadband radiation source and emits an

intense narrowband in the microwave region. Intensity and frequency of the emission

varies nonlinearly with the pumping radiation.

The big drawback is that there is a fundamental inefficiency associated with

population inversion. The photon-molecular interactions that keep the population in the

correct energy state for continual maser emission are somewhat unpredictable, which

makes it difficult to have 100% efficient use of the incoming radiation. Additionally, a

stable population conversion cycle usually contains three or more sub-transitions, which

could present a potential problem for keeping a medium in a uniform state, and producing

a narrowband emission. However, the efficiency losses described above would take the

form of thermal energy, and could therefore be possibly recycled through another energy

collection mechanism such as a heat engine.

Despite the potential problem of inefficiency there are several previous concepts

which show decent efficiencies from solar pumping. In 1963 Z.J. Kiss, H.R. Lewis, and

R.C. Duncan Jr. published a paper entitled Sun Pumped Continuous Optical Maser.24

Although this experimental efficiency is low and the device is different conceptually, it is

an example of maser system feasibility. Another more recent system concept, presented

at the Space Technology & Applications International Forum conference 2005, was Taku

Saikis Development of Solar-Pumped Lasers for Space Solar Power Station.25 It

describes a solar pumped solid state Nd/Cr:YAG ceramic laser with optical to optical

conversion efficiencies up to 38%.25 The most promising thus far is a proposal for a mid-

range infrared continuous wave laser from DARPA that cites a 50% wall plug

16
efficiency.26 These concepts will also be investigated and incorporated into later design

concepts.

2.3.7 Optical Rectenna

An optical rectenna is an antenna that couples with and rectifies optical radiation

to DC. Before ITN, in the 1960s W.C. Brown pioneered the field of wireless power

transmission with his work on microwave rectennae.27 He successfully developed a

helicopter powered by a microwave rectenna.27-28 The rectenna demonstrated efficiencies

from 85-91%.27-28 It had directive properties similar to the half-wave dipole in that there

was little variation in capture efficiency based on the angle of incidence of the incoming

radiation. His design for a light weight and flexible thin film etched circuit rectenna

weighed approximately .25 kg/m2.28

Figure 4: Diagram of W.C. Browns Thin Film Etched Circuit Rectenna27

17
 In 2002 ITN Energy Systems published its work on optical rectennae. The design

consisted of an antenna array and a high frequency metal-insulator-metal diode to rectify

the broadband field across the antenna to DC power. ITN estimated theoretical

conversion efficiencies greater than 85%. For antennae on the optical scale, ITN found

that a majority of the energy in the surface modes is carried in the dielectric above the

antenna so surface losses must be taken into consideration. Also the antenna must be

designed to couple with a complex, broadband waveform. Unfortunately only 1%

efficiency was actually achieved during testing which can be attributed at least partially

to problems associated with early stages of nanofabrication.17

Additionally, according to antenna theory, antenna length will scale linearly with

the wavelength of the incoming radiation.17 The theory is that an antenna array of a

certain scale can capture radiation of wavelengths greater than the smallest dipole

resolution of the array but no greater than the size of the whole array, much in a manner

similar to a Fourier transform. It is believed that a correctly designed and constructed

antenna array could absorb solar radiation with 100% efficiency.17 For broadband solar

radiation the antenna would be an array of tiny half-wave dipoles, corresponding to

wavelengths of approximately 2-0.3 m.17 At this scale the rectenna could also couple

with other radiation sources such as albedo encountered by planetary passes or Jupiters

radiation belts. The incident radiation induces a waveform current into the antenna array

which would be traditionally converted to DC by a non linear element such as a diode.

However, in combination with signal processing solutions and nanofabricated antennae

for transmission, the non-linear rectification elements could potentially be removed, in

18
favor of a central signal analysis and conditioning system. This concept will be further

explored in later system designs.

2.4 Conclusion of Literature Review

While many of these technologies for one reason or another are not focused upon

in the later stages of the project, many still have a fundamental aspect that lends to

possibly solving the direct conversion problem and should be kept in mind for future

exploration. These technologies will form the basis for three initial direct conversion

system concepts. These concepts include: a shocked photonic crystal system; a solar

pumped maser based on naturally occurring astronomical masers; and an optical antenna

array with central signal processing or optical Fourier Transform processing. These were

selected because it is believed that these are the analogs with the best possibility for

satisfying the project goal. The metrics for analyzing and rating these concepts are based

on satisfaction of the direct conversion criteria, and power per unit mass.

19
 CHAPTER 3

SYSTEM CONCEPTS

After the analysis of possible technologies and natural analogs, the focus of the

project has been narrowed to several potential SSP direct conversion system concepts.

These concepts include a shocked photonic crystal system, a solar pumped maser, and an

optical antenna system. These concepts will be analyzed in terms of power per unit mass.

Additionally, these concepts share several features that will be initially discussed.

3.1 General System Components

The following systems share several components that can be summarized

independently of the individual system concept discussions. These components include

solar sail parabolic reflectors and casings, along with beam focusing mechanisms.

The inclusion of a solar sail parabolic reflector is important in that it can gather a

large area of solar radiation at a relatively low mass per unit area. To get an idea of scale:

assuming an average solar energy density of 1.37 kW/m2, a 125 MW solar power satellite

at 100% efficiency would need a circular collector over 170 m in radius. A direct

conversion structure or medium this large would most likely negate any benefits of the

direct conversion itself via excessive mass. A low mass solar sail parabolic reflector

solves this problem by focusing down that same collected power onto a smaller area

which could potentially accept higher concentrations of energy. For instance, the material

has a very high reflectivity while weighting approximately 7g/m2.29 Although such a

20
material does not have high stiffness, it is not especially imperative in the absence of

gravity. Additionally, the solar sail material could be supported by distributed structural

members such as thin beams. The one exception to the implementation of a solar sail

reflector would be the nanofabricated antenna concept. This concept, as will be explained

in its corresponding section, involves a collection array that is nearly as light as the solar

sail collector and probably could not sustain high energy densities.

The concepts that require light to pass through some large cavity length will be

encased in a solar sails high reflectivity material. This material is again selected around

the same rationale. The material is light weight, with high reflectivity and doesnt need

significant stiffness. Another important feature to mention is that this casing will also

function as an unstable resonator. This is important because it increases the distance

traveled by the radiation while minimizing the overall length. Additionally, an unstable

resonator can be constructed with the correct geometry to cause the radiation to narrow

towards a resonant frequency based on the length of the cavity.

All three concepts will also be equipped with some sort of beam focusing

mechanism. Whether this is another smaller parabolic dish or a directed microwave

antenna array, it is important to have the ability to point the output of a system as a

focused beam. It is especially important in devices where scattering of particles or

crystals will be difficult to predict.

Additionally, all three concepts will be equipped with some sort of substrate and

structural framing. Considering how large most of these devices will be, it is expected

that there will be forces present that need to be mitigated. For the concepts where an

estimate was made, this was taken into account by historical percentage data obtained

21
from Space Mission Analysis and Design30 of the dry mass of a spacecraft being an

average of 3.3 times the payload mass.

3.2 Shocked Photonic Crystal System

The shocked photonic crystal system is a SSP direct conversion system based off

the work of Dr. Joannopoulos19-21 as discussed in section 2.3.5. The following sections

will describe the basic concepts, the estimation methodology, and discuss further realistic

considerations for the implementation of the system.

3.2.1 Concept Description

A diagram of the shocked photonic crystal system is shown in Figure 5. The

shocked photonic crystal system concept uses a solar sail material parabolic reflector to

focus a large collectable area of solar radiation onto the photonic crystals. The light from

the parabolic reflector is focused onto a series of shocked photonic crystals. An

alternative to a series would be one large photonic crystal with a gradient. This is

important because the functionality of a photonic crystal is based upon its internal, atomic

level, structure. There must be some sort of matching between the incident radiation

frequency and the photonic crystal. Thus in order to preserve efficiency as the band is

narrowed and the frequency downshifted, the photonic crystal will need to be changed

either serially or as a gradient. The functionality of the interface of the photonic crystal

and the solar cell will be similar to a resonator. The light will pass through the crystal

22
once and then bounce back to where it is reflected off a shock. This reflection will iterate

to the desired frequency.

The series of photonic crystals will be shielded by solar sail type material. This is

important so that there are not losses at the edge of the crystals as the light passes

throughout the possibly long length of the crystal structure. The crystals will also need to

be shocked as was an integral point to Dr. Joannopoulos simulations.19-21 As shown in

Figure 5, this requires some sort of dielectric shocking mechanism, and its own power

source. The problems associated with this will be discussed in the following sections.

Following the transmission through the crystals it is expected that the emission may need

to be focused and directed, so some sort of beam focusing dish is included in the concept.

Solar Sail Parabolic Reflector

Reflected Broadband Sunlight
Sequential Photonic Crystals
Scatter Shielding
Focused Microwave
Narrowband Emission

Beam Focusing Mechanism

Shock Power Source

Dielectric Modulation Shocking Mechanism

Figure 5: Shocked Photonic Crystal System Concept Diagram

The mechanism through which light will travel and be reflected off a moving

shock is not necessarily intuitive. Light will pass into the cavity where it will undergo

23
total internal reflection when it hits the generated shock. The Doppler shift between the

incident and reflected light due to the motion of the shock is what causes the frequency

shift. The light then bounces off the back wall of the resonating cavity again. This

process will repeated continuously until the light is shifted to the desired frequency. It is

estimated based on the frequency shift ratios from Dr. Joannopoulos work19-21 that it will

take approximately 50 shock-radiation interactions to shift the frequency from 500 THz

to 100 GHz.

3.2.2 Estimate

The estimation of such a futuristic device has significant complications. First and

most importantly is the fact that the Doppler shift in shocked photonic crystal phenomena

is at this point observed in simulation only. Realistic efficiencies for each shock-radiation

interaction, considering edge effects and nano-fabrication defects would probably be

significantly less. This would be amplified by the iterative nature of this process. Other

efficiency losses would occur in reflection at the entrance and exit interfaces of the

device, and energy lost to heat in the crystal itself.

However, if an estimate were to be made, it could be constructed around several

design constraints. First of all, it would be necessary to focus as much light onto the

smallest area of crystal as possible. This would be limited by the melting point of the

material, how much flux the crystal could accept, and how fast that radiation would pass

out. Between the reflections needed, the shock speed, density of the material, and the

thermal parameters previously specified, a estimate of the geometry and mass of the

24
cavity could be created. If the previous parameters could be defined a viable estimate

could possibly be made.

3.2.3 Implementation Analysis

In addition to the problems listed in the estimation section, there is a theoretical

problem involving the pulsed nature of this concept. It seems unlikely that all the light

will be able to be fully converted through interactions with a shock simply because the

radiation source is constant, and the shock creation and interaction is transitory.

However, perhaps a solution to this and other problems could be found, through storage

in a resonator or some other mechanism.

Despite these problems, the fact that there is a simultaneous bandwidth narrowing

and frequency shift shows that this concept is worth further study. There are several

applications where this could be vastly superior to current conventional methods.

Photonic crystals in another form could also potentially be utilized for a SSP concept in

other facets such as sharp angle wave guides20 among a whole other array of uses.

However, in comparison to other options, this concept is not as effective considering the

resulting probable high weight and effective lower efficiency. The ultimate answer is that

the further nanotechnology fabrication progresses, the closer actual efficiencies from

experimentation may approach simulated values.

25
 3.3 Solar Pumped Maser System

The solar pumped maser system is essentially an artificial analog to a naturally

occurring astronomical maser. Astronomical masers are generally caused by the

interaction of interstellar light and a low density gas cloud around a star or supernova.22-23

The following sections will describe the basic concepts, the estimation methodology, and

discuss further realistic considerations for the implementation of the system.

3.3.1 Concept Description

A diagram of the solar pumped maser is shown in Figure 6. As can be seen, the

broadband free-space solar radiation is again focused down onto the direct conversion

medium. In this case the medium is contained in a long cylindrical resonator, covered and

encased by solar sail material. This casing is very important in that it will contain the

unpredictable emissions from the radiation-molecule interactions and converge them

towards the resonant frequency of the cavity. The medium will be a low density

molecular vapor. For this concept the vapor was selected as SiO, because of the

transitions near our desired frequency output of 100 GHz, as can be seen in Table 3. This

is one of the benefits of the solar pumped maser system, in that different mediums could

be selected for applications with different desired frequencies. Again the output will also

be controlled with some sort of beam focusing and pointing dish or mirror.

26
Solar Sail Parabolic Reflector Solar Sail Material Cavity Casing

Reflected Broadband Sunlight

Focused Microwave
Narrowband Emission

Unstable Resonator Cavity Beam Focusing Mechanism

Filled with Low Density
Molecular Vapor

Figure 6: Solar pumped maser system concept diagram

Table 3: Sample astronomical maser transitions and frequency output23

Interstellar masers Stellar masers
Frequency Number Number Number Number
Molecule (MHz) known mapped known mapped
OH 1612.231 20 1 250 7
OH 1665.402 100 7 50 2
OH 1667.359 100 1 50 4
OH 1720.53 60 4 0 0
OH 4660.42 10 0 0 0
OH 4765.562 10 0 0 0
OH 6030.747 6 0 0 0
OH 6035.092 20 2 1 0
OH 13441.417 1 0 0 0
H2O 22235.08 194 19 100 5
SiO 42519.3 0 0 3 0
SiO 42802.54 1 0 20 2
SiO 43122.03 1 1 60 2
SiO 86243.35 1 0 60 0
SiO 129363.26 1 0 0 0
SiO 86846.89 0 0 1 0
CH3OH 25124.87 1 1 0 0

27
3.3.2 Estimate

One of the cruxes for a solar pumped maser system is an efficiency argument

based upon the scale of the transitions. A tested efficiency of 38%25 by a solid state laser

and a hopefully soon to be proven 50% wall plug efficiency26 implies that even higher

efficiencies are possible for a low density molecular vapor. This is due to the fact that a

laser uses electronic transitions to excite electrons through different optical modes,

whereas a maser would use molecular rotation and vibration modes of excitation. The

fact that a maser involves modes and transitions on the molecular level, whereas a laser

uses subatomic transition, should imply higher efficiency purely based on the size of the

particles dealt with. In addition the maser system would produce much longer

wavelengths, which would again simplify the system just based upon scale. The only

reason that this hasnt been proven is that there has been no driving force to develop an

efficient high frequency, high power maser. Given a fraction of the energy, time, and

money that has been expended on developing lasers, maser systems should be able to cite

higher efficiencies. For this specific case output at 129 or 86 GHz shows atmospheric

transmission efficiency at approximately 90% from Figure 1. This efficiency factor

would be multiplied to the base efficiency selected from this historical laser data

previously mentioned.

Aside from the efficiency the maser could be estimated around an energy balance.

Essentially, the molecular vapor has to be at some base energy level and pumped up to

another desired energy level. This rotation between energy levels must form a sort of

equilibrium state for the maser to work constantly over time. This equilibrium will be

28
balanced between energy flux in and out of the vapor and the required energy level of the

vapor itself, and its corresponding pressure, temperature, and density levels.

The size of the cavity was selected based on cylindrical maser models as shown in

Figure . Essentially the maser is non linear in terms of intensity up to the point of

saturation. It would therefore appear that the system would get the most performance by

operating at this boundary.

Given the proceeding information, an estimate was made for the solar pumped

maser system. The end result was a specific power of 3.54 kW/kg. This is approximately

an increase in specific power by a factor of 10.8 over a standard photovoltaic system.

This estimate was based on a standard SiO astronomical maser density of 1E10

particles/cm3 an aspect ratio of 35, and a power output of 125 MW. This puts the cavity

length at approximately 4 km. This is obviously long, but entirely in the size range of

SSP stations, yet far less massive.

29
Figure 7: Cylindrical aspect ratio relation to maser saturation23

3.3.3 Implementation Analysis

One of the most promising implementation attributes of a maser system is that

theoretically any molecule can undergo this maser transition, with nearly infinite

transition combinations between multiple molecules and energy levels. So in theory,

given further study, different molecules could be tailored or chosen to provide any

desired output frequency. The ability to provide application dependent frequencies is

extremely important to the success of the overlying concept as a whole. In addition

assuming the correct correlations and relationships could be made, it would theoretically

30
be possible to create or choose a molecule with the most preferential properties. The most

preferred properties would be dictated by properties related to the energy balance and

mass of the system. For instance, a molecule with low molecular weight may be

preferable. Or perhaps, a molecule that has dense maser emission lines at a higher energy

level would be desirable to keep the volume of the gas low by pumping more light

through a smaller area.

Despite the possibilities and benefits previously listed, and aside from the massive

resonator cavity tube length, some important implementation problems were discovered.

These include, the general high velocities in maser regions, and the general three level

requirement for optical pumping.

The high velocity problem is that often in maser regions the population inversion

occurs because particles accelerate into regions of high velocities; expanding out from a

star or a supernova.23 In such regions, the energy that was contained in random thermal

motion of the molecules is now present in directed motion, and the density also has come

down, so that collisions between molecules become rare. Under such conditions,

molecules that were excited to high energy levels in the hot regions, continue to carry

these energy levels into the cooler, fast-flowing regions, ready to give up the energy as

radiation. If this high velocity is a requirement for maser functionality, then the system

design will have to be fundamentally changed perhaps past the point of practicality. In

the simplest terms, creating and sustaining this velocity would probably take some

measure of energy which would detract from the overall efficiency of the system.

The other large problem is that masers have not been found to function from

optical pumping with under three energy levels. Now depending on the tailoring the

31
system could result in output with two frequencies very close to the desired, but for

resonation this is probably not optimal. The solution to that would be thermal pumping,

which allows two energy levels, but that would be another design issue.

3.4 Optical Antenna System

The optical antenna system is based on classical antenna theory scaled down to

optical frequencies. Whether the concept involves dispersed rectification components, or

a central processing unit, they both couple with broadband radiation through a

nanofabricated optical antenna array. The following sections will describe the basic

concepts, the estimation methodology, and discuss further realistic considerations for the

implementation of the system.

3.4.1 Concept Description

As shown in Figure , there are two possible methods involving an optical antenna

array. The first would be more similar to the conventional rectenna design and therefore

to ITNs concepts.17 However, this would be potentially more difficult considering the

need for nano-fabrication of not only the antenna array, but distributed rectification

components as well. Once the induced current is rectified, it becomes DC, and there will

be the same initial problem; that the current must be re-oscillated which requires mass,

energy, and a loss of efficiency. For the purposes of this project it would not be a true

direct conversion system.

32
 A possible solution is that the incoming broadband radiation could be filtered at

least at the highest incident frequency or preferentially even a multiple of the highest

frequency, and then additively restructured to form a lower frequency signal. It would be

effectively similar to a Fourier Approximation. This concept is roughly demonstrated in

Figure. Once formed into a lower frequency wave it could be retransmitted using a

simple microwave antenna such as is used in many current applications.

Another benefit given a high enough sampling rate and the ability to process data

at that speed would be a tunable signal. This would be extremely important when trying

to adapt the same overlying concept to various applications that have their own

application dependent frequency. Optical Fourier Transform computing components offer

possible solutions here. Also, a system with a central sampling system would entail

significantly less mass and complexity without the need for incoming wavelength scale

rectification components such as dispersed filters and diodes. This could potentially

increase both efficiency and specific power.

33
 Conventional Newer Optical
Optical Rectenna Antenna with Central
Concept Signal Processing
Concept
Central Signal
Processing Unit

Gyrotron or Similar
Nanofabricated Broadband Optical Transmission Device Less Dense
Rectenna with Distributed Rectification Nanofabricated
Circuitry Broadband Optical
Antenna Array
(Similar to ITN and W.C Brown)16,26-27
Microwave Antenna
Figure 8: Optical antenna and rectenna concept diagrams

Low Frequency Wave Approximation with High

Frequency Broad Band Sampling
1.5
Target f=f0
1
f=8*f0

0.5 f=4*f0
Amplitude

f=2*f0
0
f=f0
0 10 20 30 40 50 60
-0.5 f=3*f0

Approximation
-1

-1.5
Time

Figure 9: Simulation of low frequency wave approximation by 2x max frequency

sampling

34
3.4.2 Estimate

Simply based on of cited numbers from W.C. Brown and ITN,16,26-27 an estimate

for a SSP system incorporating an optical antenna can be made. ITN cited a range of

theoretical coupling efficiencies of 85-100%. Taking the low end of this range, using

W.C. Browns estimated weight for a thin film etched circuit rectenna, and assuming an

average solar energy density of 1.37 kW/m2 gives the following attributes: a power

density of 1.165 kW/m2 and a specific power of 4.658 kW/kg.

Comparing against the upper end of the range of data cited in Table 1 for

photovoltaic specific power shows that an optical rectenna could provide a 1763%

increase in specific power over conventional photovoltaics.

It must be noted that this estimate, like all other estimates in this paper, has a large

uncertainty that highly depends on which end of the range is used. Because of this,

finding a specific uncertainty value would be unproductive. However, the estimate can

still serve as an example as to the providence of the concept. In addition these estimates

generally take values at the ends of the ranges in favor of traditional photovoltaics.

3.4.3 Implementation Analysis

It is also important to note that the optical rectennae should still be seriously

considered with respect to conventional space solar power without the need for THz

sampling. Just based off the specific power improvement, if SSP concepts utilizing

35
microwave generators had their standard photovoltaics replaced with optical rectennae,

the system might be improved to the point where it is actually feasible.

Optical rectennae could possibly also be applied to more futuristic concepts. For

instance in the case of a problem with direct sunlight pumipng, an optical antenna grid

could be used to pump a laser for re-transmission using the broadband signal. Despite this

possibility, the reasoning was made that if any laser or maser type device could be

pumped by a broadband signal from an antenna array, then it should still be able to be as

easily pumped directly by sunlight focused by a parabolic reflector. This reflector could

be built out of solar sail type material and would be theoretically lighter and significantly

easier to construct than a nanofabricated optical antenna.

In general the optical antenna or rectenna array appears to be the most promising

direct conversion concept. Nanofabrication is advancing rapidly and actual functional

tests may be possible in the near future.

3.5 Conclusion of System Concepts

Three futuristic SSP direct conversion system concepts were explored. Following

the analysis an estimate for specific power was made where possible, and an underlying

decision as to the feasibility of the concept was made. These estimates and decisions,

along with the corresponding rational are shown in Table 4.

36
Table 4: Summary of direct conversion system concepts specific power estimates
and conclusions

Specific
Power
Concept (kW/kg) Decision Rationale
Shocked Photonic Not enough Unresolved Won't realistically see these
Crystal data Issues efficiencies, difficulty with
pulsed functionality and shock
creation
Solar Pumped Laser 3.34 Unresolved Difficulty in the creation of an
Issues equilibrium of efficient
transitions at desired
frequencies
Optical Antenna 4.66 Possible Need nano-fabrication and or
THz wave sampling
Conventional 0.31 Possible Technology is there, but may
Photovoltaics not be economically feasible

37
 CHAPTER 4

SELECTED APPLICATIONS

Applications for which direct conversion for space solar power would be useful

are plentiful. Essentially any architecture that needs power would possibly serve to

benefit from direct conversion. Two major architecture categories that are the most

obvious benefactors of direct conversion are space-based power grids, and electric

propulsion. It is important and interesting to note, that direct conversion would enable

even further possibilities, such as a combination of the above. For instance, space power

grids and electric propulsion could possibly be used together to create a network of power

supply to enable expansion throughout the solar system and beyond.

4.1 Space Solar Power Grid

Aside from saving the world, space solar power is a great idea that has been

around for a fairly long time. It ultimately offers nearly free and unlimited energy. In

response to skeptical comments as to a space based solar power grid, further explanation

as to the benefits of a SSP grid versus ground based solar power system will be reiterated.

Primarily the SSP system gets 24 hours per day sunlight versus a terrestrial maximum of

8, resulting in an initial factor of three times the generating capacity. Additionally, there

are atmospheric effects as that drastically reduce the solar energy density at the ground by

30-40%. Ground based solar power is also affected by seasons, weather, and land

availability. Functional SSP would be a tremendous breakthrough. The main issue faced

38
by such a project is the state of current technology. Specifically, launch costs, solar cell

efficiency, and an infrastructure that would be receptive to such a grid.

Based on the estimates listed in Table 5, the benefit of alternative SSP

technologies is evident. Cost savings of this magnitude transform any SSP concept into

feasibility. Systems designed around 100 GHz would significantly reduce ground based

infrastructure and therefore cost by limiting beam spreading. Compared to a commonly

cited 2.4 GHz system the ground foot print of the beam should be nominally 2.4% as

large.

Cost comparisons were created through the average values of photovoltaic

specific power and launch costs from Table 1 and Table 2 respectively. The specific

power values for the direct conversion options were calculated as described in the

proceeding sections and shown in Table 4.

Table 5: Cost comparison of direct conversion concept implementation on a 125

MW solar power satellite

125 MW Output SSP Specific Power Mass Cost Cost of PV

Satellite (kW/kg) (MT) ($M) (%)
Conventional 0.31 403.23 12741.94 100.00
Photovoltaics
Nanofabricated Optical 4.66 26.82 847.64 6.65
Antenna
Solar Pumped Maser 3.34 37.43 1182.63 9.28

The SSP system most preferred by NASA to-date is based on placing very large

photovoltaic arrays in geosynchronous Earth orbit, which are positioned to beam power

down to large ground stations around the equator. The primary problem with this

approach is that until the system in GEO is operational, no revenue can be generated. The

initial cost is extremely large. Even at an unrealistic estimate of $100 per pound launch

39
cost to geosynchronous Earth orbit, the cost to first power was estimated at $300 billion.

With launch costs today, as shown in Table 2, it is easy to see why SSP has not been

implemented.

With the savings cited in Table 5, combined with a revenue-generating

evolutionary path to SSP could solve the chicken and the egg issue of developing such a

system. Prof. Komerath has detailed such an evolutionary path in several papers.5-7 The

evolutionary path is to develop a space based power transmission system, which would

pay for itself from the savings from transmission costs. This would be accomplished by

allowing ground-based renewable energy plants to trade their output to where it fetches

the best prices, thus enabling the development of new renewable energy plants at better

locations. These locations are typically far from metropolitan areas and industrial centers.

Additionally, the price differential between off-peak and peak time power costs is more

than adequate to offset the lower transmission efficiency of initial SSP systems, so that

the SSP grid could pay for itself over a 10-15 year period. Further, such a system would

enable green energy sources such as wind and solar plants to become base load sources

and exchange power anywhere across the world.5-7

4.2 Electric Propulsion

Over the years a multitude of electric propulsion systems have been proposed and

developed. The major drawback of these systems is a high mass per unit thrust, due to the

power source and transmission system. One of the most commonly known forms of

electric propulsion, ion engines, work by ionizing propellant particles by EM radiation

40
such as xenon gas and accelerating them through an electric field.31 Magnetoplasma

engines are a newer concept but work on somewhat similar principles: heating neutral

hydrogen gas into plasma using electric fields and contained by magnetic fields, the

plasma then passes through an RF booster to further ionize the hydrogen plasma.32 The

University of Washington has also been looking into various other futuristic electric

plasma based propulsion methods such as the MagBeam.33

These propulsion systems could all be revolutionized by the massive amounts of

energy available from the sun that could be harnessed by direct conversion systems.

Energy could also be beamed from a solar power grid. This would eliminate one of the

important opposition points to electric propulsion; that electric propulsion must often

utilize massive on board power systems such as nuclear power generation in order keep

power over the long range in which electric propulsion becomes effective. Direct

conversion options eliminate the need for an onboard power system, which would both

decrease the launch cost and increase the effectiveness of the propulsion system.

Furthermore, direct frequency conversion is especially applicable because of the

need for EM radiation to ionize particles. Mass and efficiency could be saved by directly

converting to the needed radiation frequency. These mass savings are extremely

important because of the low thrust nature of electric propulsion systems.

As specified by the designs the direct conversion systems all either utilize a solar

sail reflector or a thin film array. In the case of the thin film array, perhaps with future

advances in fabrication technology, direct conversion technology might be combined

with a solar sail. Landis suggests a hybrid electric propulsion and solar sail system.33 In

41
addition to direct conversion applying to the electric propulsion system, it would help as

a sail by providing initial thrust and the magnitude of thrust needed for a mission.

4.3 Conclusion of Selected Applications

It is evident that between the combination of pure specific power increase and

integration advantages that direct conversion systems provide great benefits to many

space based architectures. Although based on rough estimates, possible mass savings of

this magnitude warrant further exploration and consideration.

42
 CHAPTER 5

CONCLUSION

This project studied a large number of options that might lead to direct

conversion. Those technology options were analyzed according to which would warrant

further exploration from the point of view of aerospace systems applications and possible

power per unit mass. Based off these technologies, several advanced concepts were

proposed. Estimates of the possible power per unit mass of potential direct conversion

systems were made, and future applications that would benefit from those direct

conversion systems were identified.

The project goal, concepts developed, critical enabling technologies, and

applications are detailed in Table 6.

43
Table 6: Direct Conversion for SSP summary table
Direct Conversion for SSP Summary:

Goal: Convert broadband solar radiation directly to lower narrowband frequency

Concepts:
1. Optical Rectenna System: Solar radiation is coupled to thin film
nanofabricated antenna array including either rectification components or
optical-transform components for THz wave reconstruction
a. Efficiency: 85-100%
b. Specific Power Estimate: 4.658 kW/kg
2. Shocked Photonic Crystal System: Solar radiation gathered by solar sail-like
parabolic reflector is focused through tailored dielectrically modulated shocked
photonic crystal which simultaneously narrows the band and downshifts the
frequency
3. Solar Pumped Maser System: Solar radiation gathered by solar sail-like
parabolic reflector is focused through a low density molecular vapor contained
in a cylindrical resonator tube constructed from solar sail material producing a
non-linear emission
a. Efficiency: > 50%
b. Specific Power Estimate: 3.539 kW/kg

Critical Technologies:
1. Nanofabrication: sub 100 nm resolution
2. Control System: utilizes incoming power to stabilize SSP systems in orbit
3. Ultra Thin Solar Sail Fabrication
4. Terahertz Optical Computing/Transform Elements: > 3000 THz

Applications:
1. Space Solar Power Grid
2. Electric Propulsion

To reiterate, direct conversion offers the potential to realize SSP as a feasible

aerospace concept. The ability to convert broadband sunlight directly to a useable

narrowband application dependent frequency reduces system mass and complexity while

increasing efficiency and specific power. Out of the diverse array of possible

technologies and options studied, a handful of concepts have shown potential to warrant

further study. They have the potential to be enabling technologies in their own right for

SSP systems and be a true paradigm shift for SSP.

44
 REFERENCES
1
Mankins, J. C., "A fresh look at space solar power: New architectures, concepts and
technologies." Acta Astronautica 41(4-10): 347-359. 1997.
2
Glaser, P. E., "Power from the sun: Its future." Science, 1968 Vol. 162, pp. 857-861.
3
Itoh, K., Ogawa, Y., Omiya, M., "Project overview and prospect of solar power
satellite, SPS2000." Hokkaido Daigaku Kogakubu Kenkyu Hokoku. Bulletin of the
Faculty of Engineering, 1995, Hokkaido University Vol. 175, pp.113.
4
Komerath, N., Boechler, N., Wanis, S., Space Power Grid- Evolutionary Approach To
Space Solar Power. Proceedings of the ASCE Space and Earth 2006 Conference,
League City, TX, April 2006.
5
Boechler, N., Hameer, S., Wanis, S., Komerath, N.M., An Evolutionary Model for
Space Solar Power. In El Genk, Editor, STAIF 05-082, Proceedings of the Space
Technology and Applications International Forum, American Institute of Physics
Conference Proceedings Volume 813, Albuquerque, NM, Feb. 2006, ISBN: 0-7354-
0305-8.
6
Komerath, N.M., Boechler, N., The Space Power Grid. Paper IAC06-D3.4.06,
Proceedings of the 56th International Astronautical Congress, Valencia, Spain, October
2006.
7
U.S. Department of Energy, Bandgap Energies of Semiconductors and Light. Dec
2005. Retrieved 4/14/2007. http://www.eere.energy.gov/solar/bandgap_energies.html
8
Fatemi, N. S., Pollard, Howard E., Hou, Hong Q., and Sharps, Paul R. (2000). Solar
Array Trades Between Very-High Efficiency
Multi-Junction and Si Space Solar Cells. 28th IEEE PVSC. Anchorage, Alaska.
9
Spectrolab Inc., "Spectrolab Photovoltaic Products Data Sheet." Oct. 2004., Retrieved
4/14/2007, from http://www.spectrolab.com/DataSheets/Panel/panels.pdf.
10
Murphy, D. M., Eskenazi, M. I., White, S. F., Spence, B. R., Thin-film and crystalline
solar cell array system performance comparisons. AEC-Able (ABLE) Engineering. New
Orleans, LA, United States, Institute of Electrical and Electronics Engineers Inc. 2002.
11
Olds, J., Way, D., Charania, A., Budianto, I., Marcus, L., "In-Space Deployment
Options for Large Space Solar Power Satellites," IAA-00-R.2.02, 51st International
Astronautical Congress, Rio de Janeiro, Brazil, October 2-6, 2000.
12
Isakowitz, S. J., Hopkins, J. B., Hopkins, J. P., International reference guide to space
launch systems. American Institute of Aeronautics and Astronautics, 2004. Reston,
Virginia.

45
13
Wade, M., Encyclopedia Astronautica. 1997-2007, Retrieved 4/14/2007, from
http://www.astronautix.com/
14
Robel, Michael K. The cost of medium lift. The Space Review. June 1, 2004.
http://www.thespacereview.com/article/150/1
15
ARO (Arizona Radiowave Observatory) Website: What is Submillimeter
Astronomy? http://kp12m.as.arizona.edu/docs/what_is_submillimeter.htm Accessed
5/1/2006.
16
Ren, A., Kuo, S.P., Frequency Downshift in Rapidly Ionizing Media. IEEE,
Piscataway, NJ, USA. 1994.
17
Berland, B., PhotoVoltaic Technologies Beyond the Horizon: Optical Rectenna Solar
Cell. Final Report, NREL/SR-520-33263, February 2003.
18
Iltchenko, V., Matsko, A., Savchenkov, A., Maleki, L., A Resonator for Low-
Threshold Frequency Conversion. JPL.
http://www.nasatech.com/Briefs/Dec04/NPO30638.html
19
Joannopoulos, John D., Reed, E., Soljacic, M., Reversed Doppler Effect in Photonic
Crystals. Physical Review Letters. Sept 2003.
20
Joannopoulos, John D., Johnson, Steven G., Photonic Cystals: The Road from Theory
to Practice. Massachusetts Institute of Technology. 2002.
21
Joannopoulos, John D., Reed, E., Soljacic, M., Color of Shock Waves in Photonic
Crystals. Physical Review Letters. 23 May 2003.
22
M. Reid, J. Moran, Masers. Annual Review of Astronomy and Astrophysics 1981.
23
Elitzur, M. Astronomical Masers. Kluwer Academic Publishers. 1992.
24
Kiss, Z. J., Lewis, H. R., Duncan, R. C. Jr., Sun Pumped Continuous Optical Maser.
Applied Physics Letters. March 1963.
25
Saiki, T., Uchida, S., Motokoshi, S., Imasaki, K., Nakatsuka, M., Nagayama, H., Saito,
Y., Niino, M., Mori, M., Development of Solar-Pumped Lasers for Space Solar Power
Station. Space Technology Applications International Forum. October 2005.
26
DARPA Efficient Mid-wave Infrared Lasers (EMIL) BAA06-20.
http://www.fbo.gov/spg/ODA/DARPA/CMO/BAA06%2D20/Attachments.html
27
Brown, W.C., The History of Power Transmission By Radio Waves. IEEE
Trans.Vol. MTT-32, p:1230 (1984).

46
28
Brown, W.C., Performance characteristics of the thin-film, etched-circuit rectenna.
IEEE MTT-S International Microwave Symposium Digest, p: 365- 367 (1984).
29
Solar Sail Technology Development: Mission Senarios JPL. Mar 2002.
http://solarsails.jpl.nasa.gov/introduction/mission-scenarios.html
30
Wiley L.J., Wertz J.R., Space Mission Analysis and Design, 3rd Edition Microcosm
Press; 3rd edition (October 1999)
31
NASA's Human Exploration and Development of Space Enterprise, Propulsion
Systems of the Future. 15 May 2003.
http://www.nasa.gov/vision/space/travelinginspace/future_propulsion.html
32
Winglee, R., Magnetized Beamed Plasma Propulsion (MagBeam). NIAC. March
2005.
33
Landis, Geoffrey A., Optics and Materials Considerations for a Laser-propelled
Lightsail. Paper IAA-89-664 at the 40th International Astronautical Federation
Congress, Mlaga, Spain, Oct. 7-12, 1989. Revised December, 1989.

47