On the challenge of a century lifespan satellite

Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at ScienceDirect

Progress in Aerospace Sciences journal homepage: www.elsevier.com/locate/paerosci

On the challenge of a century lifespan satellite Jesús Gonzalo a,n, Diego Domínguez b, Deibi López a a b

University of León, Spain Vrije Universiteit Brussel, Belgium

art ic l e i nf o

a b s t r a c t

Article history: Received 24 February 2014 Received in revised form 28 April 2014 Accepted 8 May 2014

This paper provides a review of the main issues affecting satellite survivability, including a discussion on the technologies to mitigate the risks and to enhance system reliability. The feasibility of a 100-year lifespan space mission is taken as the guiding thread for the discussion. Such a mission, defined with a few preliminary requirements, could be used to deliver messages to our descendants regardless of the on-ground contingencies. After the analysis of the main threats for long endurance in space, including radiation, debris and micrometeoroids, atmospheric drag and thermal environment, the available solutions are investigated. A trade-off study analyses orbital profiles from the point of view of radiation, thermal stability and decay rate, providing best locations to maximize lifespan. Special attention is also paid to on-board power, in terms of energy harvesting and accumulation, highlighting the limitations of current assets, i.e. solar panels and batteries, and revealing possible future solutions. Furthermore, the review includes electronics, non-volatile memories and communication elements, which need extra hardening against radiation and thermal cycling if extra-long endurance is required. As a result of the analysis, a century-lifetime mission is depicted by putting together all the reviewed concepts. The satellite, equipped with reliability enhanced elements and system-level solutions such as smart hibernation policies, could provide limited but still useful performance after a 100-year flight. & 2014 Elsevier Ltd. All rights reserved.

Keywords: Lifespan Spacecraft design Endurance Reliability Survivability Space weather

Contents 1. 2.

3.

4.

5.

n

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Analysis of lifespan killers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2.1. Upper atmosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2.2. Thermal environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.3. Radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.4. Debris . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.5. Global effects of aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Strategies and available technologies to enhance survivability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3.1. Satellite orbit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3.2. Size, shape and shielding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3.3. Fault tolerance: redundancy and reliability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 3.4. Power sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3.5. Onboard power storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.6. Data communications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Discussion on long lifespan satellite concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 4.1. Flat plate shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 4.2. Tetra-decahedron shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Corresponding author. Tel: +34 987 291000x5251. E-mail address: [email protected] (J. Gonzalo).

http://dx.doi.org/10.1016/j.paerosci.2014.05.001 0376-0421/& 2014 Elsevier Ltd. All rights reserved.

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

2

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

1. Introduction Lifespan is a requirement of space missions with an obvious but complex impact on the cost and the revenue [55]. As hands-on maintenance of space systems is almost impossible, lifespan is driven by the reliability of the elements and their architectural integration to achieve the final objective. From a practical point of view, reliable systems are more expensive in terms of materials, components, manufacturing and testing, but provide more robust business models [54], lower insurance prices and more time to recover the investment. The present paper reviews the feasibility drivers of a mission with an enhanced lifespan of 100 years. If successful, the satellite could be able to transport a data load into the future and deliver it to our descendants, regardless the existence of the companies or individuals that created the mission. This extended lifespan poses stringent requirements on all the subsystems on-board and on ground [56]. However, since there are running proves working for the last four decades [30], Earth observation satellites like Landsat for 30 years [35,16], other exceeding many times their expected lifetime [68] and new communication satellites declaring nominal figures of around 20-year lifetime [11], the challenge seems to be achievable in the near future. The review below is focused on the space segment, because the proposed long-lasting mission should be autonomous enough to survive without specific contact from ground. Conventional missions are built on a space–ground interaction scheme. The unique characteristic of the new concept is the fact that the ground segment built for the satellite is not essential for the success of the mission. Thus, during the first months/years, a ground facility can be used to monitor the satellite and to complete the data upload. However, at some point in time, the satellite must make its way on its own, in a deep hibernation flight, until the time of data download is reached. At that point, the satellite should be prepared to communicate to a generic ground segment, requiring a minimum level of hardware and protocol knowledge. These points also deserve an analysis from the endurance point of view. Thus, the study includes analysis and trade-off's of orbital profiles, radiation hardening, thermal protection, energy harvesting and accumulation, data storage and communications. All of these are combined with reliability enhanced mechanisms such as homogeneous and heterogeneous redundancy and system-level solutions such as smart hibernation policies. The analysis of the available technologies and the proposed solutions permit the development of the first satellite concepts for this revolutionary achievement, which would push the terms lifespan and autonomy further than what is envisaged today.

Although the economic issues are not within the scope of this work, all the decisions suggested during the concept definition look for affordable solutions, presently available or expected in the very near future. As a consequence of that, precursor satellites meeting the centennial lifespan could be launched in a few years from now. The space environment strongly influences the performance and lifespan of any space system. This is a major issue for satellite developers, especially in this case where the expected operational lifespan is as long as 100 years. Vacuum, exposure to extreme temperatures, debris, meteoroids or radiation configure a very hostile environment threatening satellite survivability [38,64]. Thus, a good understanding of the space environment and its impact on individual elements or components on-board is needed in order to define the constraints and to choose the best technological solutions from the endurance point of view. Electronics, communications and power system malfunctions are typical causes of mission termination [66]. With respect to the on-board digital storage and data processing, which are also prone to radiation effects [63], it is necessary to fix some top level requirements in order to focus the discussion on representative applications. In this paper, the target mission would be required to carry 24 GB of information and to be able to download those data in the last phase of the mission. Besides, in order to meet the lifespan requirement, a dedicated power management policy needs to be established. Thus, the satellite should be hibernating most of the time, with all but a small waking-up device disconnected. This will limit capabilities that are present in normal satellites such as battery charging, data processing, ground communications, attitude stabilization and others along most of the mission. The satellite would be able to wake up from time to time to check data integrity and, if required, to perform some data update. In this sense, the Rosetta comet-hunter mission [24,23] is a definitive proof of the technique to achieve long missions using hibernation periods. Finally, after the 100-year trip, the satellite should provide its full functionality during a limited time. In order to allow some preliminary estimations on the system feasibility, some budgetary issues need to be taken into account. Thus, three operational stages could be considered, so-called Trip, Hello and Duty. The first stage is 100-year long, with quarterly short contacts with ground; the second is a period of maximum 1 month in which the satellite tries to catch the attention of a ground receiver using an emitting beacon; the third is the effective time during which the payload is downloaded, pass after pass, to one or more ground facilities. Taking into account typical power consumptions of conventional transmitters and the timing of the contacts, Table 1 provides reasonable power and communications budgets for all the stages of a reference mission. These figures will help with the discussion for the rest of the paper.

Table 1 Reference budgets used along the paper for the discussion on lifespan extension.

2. Analysis of lifespan killers

Av. power (W) Bandwidth (kbps) Operation time per contact (min) Contact frequency Stage duration Operation time (h) Total energy (W h)

Trip stage

Hello stage

Duty stage

2 9.6 10 4/year 100 years 67 133

1/30 1.2 1 1/h 10 days 2 24

10 2048 10 3/day 32 days 25 267

Total

117 424

2.1. Upper atmosphere Although the satellite environment can be considered as a vacuum for many applications, this is not the case when dealing with low Earth orbits, especially if lifetime is a concern. The atmosphere limit is not clearly defined but it is possible to find gas molecules, primarily oxygen in the range of 80–90 km to 500 km height, and hydrogen and helium beyond 500 km [42]. The interaction between these

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

particles and the surface of the satellite decreases its speed, driving it to slightly lower orbits gradually, in a process known as orbital decay. When the orbit is too low, heat exchange and particle impacts finally disintegrate the spacecraft. The average effect of the satellite–atmosphere interaction mainly depends on three factors [43]. The first one is the neutral density of the upper atmosphere, which is the major source of error in modelling. Nowadays, two widely used models can describe the temperature and density up to the exosphere [64]: the NRL Mass Spectrometer, the Incoherent Scatter Radar Extended Model NRLMSISE‐00 [48] and the Jacchia‐Bowman 2006 JB‐2006 model [10]. The last one is more accurate but it does not include altitudes of less than 120-km height. The second factor is the relative velocity, since the assumption that atmosphere rotates with the Earth may induce some uncertainty due to the presence of strong winds which are currently very hard to predict [27]. Finally, the shape of the satellite is modelled through the drag coefficient (around 2.2 in most of LEO cases) [43]. Modern computers can use variable drag coefficients, calculated taking into consideration the satellite attitude. The drag coefficient is a fundamental part of the ballistic coefficient, together with the spacecraft mass and cross section area, which drives the satellite deceleration and hence the orbit decay rate. There are several approaches to obtain lift and drag coefficients. For simple shapes, analytical expression can be obtained by two different main approaches of gas–surface interaction mathematical model (GSIM), Schamberg [59] and Schaaf and Chambre [58]. In the case of more complex geometries, these coefficients can be estimated by numerical methods, such as Panel Methods [49], Test-Particle Monte Carlo [17] or Direct Simulation Monte Carlo [7]. On top of that, the density is influenced not only by height but also by the solar activity [74], mainly because of UV radiation emitted by the Sun. The periods of maximum solar radiation are harmful for the lifetime of the satellite, since in these periods the density in the range 500–800 km height are two orders of magnitude higher than in minimum solar activity ones [48], and therefore the decay is faster. Generally speaking, higher initial altitudes and better ballistic coefficients ensure longer lifetimes. As an example, a 400-kg satellite with a ballistic coefficient around 0.005 m2/kg at 600 km initial orbit has an estimated lifetime of around 40 years, whereas a 1 kg cubesat with a ballistic coefficient of 0.02 m2/kg in the same initial orbit, would decay in a temporal range between 4 and 7 years, depending on the solar cycle. One last issue related to upper atmosphere environment is the large kinetic energy of the particles hitting the spacecraft (e.g. an oxygen atom at 7.8 km/s is about 5 eV, enough to break molecular bonds in the order of 1–2 eV). Therefore, these impacts can be highly reactive, presenting a destructive potential for a careless spacecraft. 2.2. Thermal environment Every component in an orbiting spacecraft has its own temperature as a result of several heat interactions. There are inputs from the Sun and from other satellite elements (radiation and/or conduction) and, for low orbits, free molecular heating. There are outputs such as the heat evacuated to deep space or to other colder satellite components. The direction of heat exchanged with Earth depends on the satellite and Earth temperatures and the sunlight albedo. There is also self-generated heat if an element has active power consumption. Furthermore, all these interactions are dynamic, as the spacecraft is coming into sunlight-eclipse cycles, the Sun is fluctuating seasonally and yearly and finally the operational requirements involve variable on-board power consumption. In an extremely long mission as the one proposed, the stability of the orbital configuration, given by the satellite mechanics coupled with the operational concept, needs to be ensured.

3

Thermal control is a discipline as old as space exploration. The above-explained heat fluxes can lead to strong differences of temperature among regions of the same satellite, e.g. the side facing the Sun and others facing cold space [57]. Needless to say that there exist relatively simple passive mechanisms to deal with many of the thermal problems, and also more elaborated ones, typically used in large satellites to provide special thermal environments for delicate pieces of equipment like sensors or antennae. When reliability is the issue, monolithic solutions such as selective coatings (different absorption–emission figures for the visible and infrared wavelengths), multi-layer insulation and blanket barriers are preferred. Furthermore, conducting structures may help to homogenize temperatures inside the vehicle. The selected materials (through their specific heat capacity) and the masses involved drive the inertia of the system to thermal changes. The thermal environment has an important impact on the aging of the basic components of the satellite, those being mainly structural and electronical [53] and their packaging [75]. These are normally susceptible to thermal cycling degradation and eventual damage. From experience (in-flight measurements as well as laboratory tests), the solar absorptivity αs is known to be the most significant parameter affected by in-orbit aging, usually increasing with time [50]. An important gain in the value of αs could significantly raise the on-board temperatures at the end of the mission. In order to compensate this effect, the satellite could require oversized radiative surfaces that, on the contrary, would involve thermal problems at the beginning of satellite life. The change in αs is mainly due to the interaction of the thermal coating with solar UV radiation and, for LEO, the atomic oxygen [51] present in the upper atmosphere. The effect is highly dependant on the choice of material: silver and aluminium offer better stability (αs increased about þ0.1 after a 7-year exposure in LEO) while white paints change their radiometric performance more markedly (αs increased by þ0.2 in the same conditions). On the other hand, the orbital environment is capable of producing significant temperature variations as the spacecraft passes from sunlight to shadow. The number and length of thermal cycles during a mission is extremely dependent on the orbital profile, from the typical 15/16 a day for most of the LEO to the daily frequency for GEO. There are no-eclipse conditions for LEO dusk–dawn configurations or GEO in summer/winter seasons. However, there are cases/missions where rare moon eclipses may need to be considered. Besides, the amplitude of the thermal cycle experienced by a certain material usually depends upon its solar absorptivity and thermal emittance, the view factors to the Earth and the Sun, the eclipse durations and the internal thermal connections or insulations. Such thermal cycles can suppose a threat for the survivability of the spacecraft for several reasons [19]. First, nonhomogeneous materials in close contact, when having different thermal expansion coefficients, may cause cracking or delamination. Second, mechanical properties of the materials (especially for polymers) can notably change with the temperature, in such a way that they could offer a decreased strength or ductility during part of the orbit. Third, radiation can alter material properties, mainly close to the surface, making them inhomogeneous through the thickness and more susceptible to suffer from a mismatch in the thermal expansion coefficient. The thermal cycling has a negative impact on composite materials and paint coatings. Carbon fibre/epoxy composites under vacuum and thermal cycling conditions experience microcracking, reaching a saturation level after a number of cycles. The depth of the thermal cycle (delta of maximum and minimum temperatures) emphasizes its effect. Some parameters, like compressive strength, are notably reduced up to certain limits (e.g. 15%) [46]. Referring to paint coatings, microcracking has been observed as a result of the difference in thermal expansion coefficient between the substrate

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

4

and the coating. A proper substrate preparation, as well as the use of anodized aluminium substrates, can minimize the degree of microcracking and improve the durability of the painted surface [19]. 2.3. Radiation A major factor affecting spacecraft reliability, especially electronic components, is radiation. Radiation is normally referred to as the transfer of energy by means of particles, including protons, in contrast to electromagnetic waves below X-ray band such as UV, visible, thermal or microwaves [63]. There are three sources of radiation in the Earth orbit: trapped radiation, Solar Particle Events (SPE) and Galactic Cosmic Rays (GCR). Trapped radiation mainly consists of electrons and ions, with energies up to 400 MeV, trapped by the Earth magnetic field in toroid regions known as Van Allen radiation belts. The SPE occurs in association with solar flares and is characterized by a rapid increase in the flux of energetic particles, up to 1 GeV, with a typical duration between several hours and several days. Finally, GCR are particles reaching the Earth from outside the solar system, composed mainly of protons but also heavy ions with energies over 200 GeV. The global fluxes of GCR are much lower than the others, although their very high energy makes them quite dangerous, especially for the electronic devices. In general, lower orbits avoiding polar regions are more protected from radiation, although the South Atlantic Anomaly can increase the expected doses. Damages produced by radiation hitting electronic or structural devices can be grouped into two different categories: cumulative effects of the dose received, known as total ionizing or nonionizing doses (TID and TNID respectively), and effects of a single particle hitting the device, named single event effects, mainly originated by heavy ions and protons. TID may produce noise and current leakage, as the radiation adds change in the dielectrics and conductors. TNID modifies the crystal lattice by displacing atoms, which may produce important alterations in component behaviour depending on the internal structure. Single event types are numerous, ranging from recoverable effects like memory flips to destructive structure ruptures. A good description of radiation sources and effects is provided in [62]. Current developments in spaceborne electronics allow for the hardening of electronics by means of passive shields or active fault tolerance [33], at the expense of mass in the first case and throughput overhead and power in the second case. 2.4. Debris Space debris – consisting of interplanetary dust, meteoroids and artificially created parts – is another significant environmental issue. The number of debris in LEO with sizes of 1 cm or greater exceeds 300,000; part of this debris (size o5 cm) is untraceable from Earth and, as a consequence, unavoidable [45]. Although this number sounds enormous, extra-terrestrial space is actually relatively empty. The chance of collision depends on the density of the traffic and on the trajectories of the objects. Thus, when trying to determine the risk for a certain satellite, it is important to distinguish between the characteristics of an orbit and the location and state of motion of a satellite in that orbit [25]. A higher risk of collision is found in the most heavily populated paths: ones where the satellite's orbit passes over the same spot on the ground at the same time every day (Sun synchronous) [73] and those where the satellite has an orbital period equal to the Earth's rotational period (geostationary) [2]. The uncertainty when predicting the evolution of debris population is certainly large. The amount of artificial debris in orbit will be strongly influenced by future human practices in space access. In the framework of an international effort to

diminish as much as possible the amount of new debris in orbit, the Inter-Agency Space Debris Coordination Committee (IADC) has published debris mitigation guidelines, with an emphasis on cost effectiveness, that can be considered during the planning and design of spacecraft and launch [32]. Latest efforts focused on forecasting the stability of the future LEO environment suggest that, even with a 90% compliance of the commonly-adopted mitigation measures for new launches, the LEO debris population is expected to increase by an average of 15% in the next 100 years [31]. An even much worse and worrying situation is predicted when a Business as Usual situation (no mitigation measures are taken) is considered. In order to assess the compliance of new spacecraft designs with the space debris requirements, two models and their corresponding software tools are commonly used: ESPENVIS [22] from the European Space Agency and DAS [44] from NASA. Because of the increasing risk of collision, spacecraft operators are paying special efforts to the development of operational procedures for debris avoidance [29]. Collision risk analysis is usually automated by daily monitoring of close conjunctions within spacecrafts under the responsibility of the operator and all the objects included by the United States Strategic Command (USSTRATCOM) in the Two-Line Element (TLE) catalogue. In the near future, the Space Surveillance and Tracking Segment, belonging to the European Space Situational Awareness Program, will also provide this data. Once dangerous conjunctions are corroborated, avoidance manoeuvre is then calculated taking into account satellite mission and platform constraints, as well as new high-risk conjunction events in the post-manoeuvre trajectory [39]. Usually, collision risk assessment relies on an exclusion volume or a probability threshold method, although other criterions like one defined by an adaptive exclusion area have been also proposed [47]. These procedures, although highly reliable for the avoidance of catastrophic collisions, are only limited to satellites equipped with propulsion subsystems, which makes them heavier and more expensive.

2.5. Global effects of aging Although small satellites are in some way prone to infant mortality [20], once the first stage of the mission has been successfully executed, the decrease in reliability due to wear-out is less dramatic than the one observed in larger satellites. This is perhaps attributable to the simplified architecture of small spacecraft. The above-mentioned study, based on more than 1500 real missions between 1990 and 2008, analyses how well a 2-Weibull mixture distribution matches the available data for satellite failure times. With such an approach, the first Weibull shape parameter (β1 o1) captures satellite infant mortality, where failure rates are decreasing with time, whereas the second shape parameter (β2 41) captures satellite wear-out failures, where failure rates increase with time. An extra parameter weights both distributions one against the other [13]. The given coefficients, taking into account that the mission envisaged in the present study is in the heavy rugged end, are useful to extrapolate reliability figures into the future, e.g. 100 years (Fig. 1). An estimated 100-year end-of-life reliability of almost 80% is shown, which is a reasonable value for a single satellite after such a long period. Furthermore, wear-out effects considered so far relate only to pieces of equipment under continuous (or at least frequent) operation. A mostly hibernating satellite, although still suffering radiation bombing and thermal cycling, should diminish chances of failures due to active component malfunction. Should the investment require higher confidence levels, redundant satellites could be considered; for the sake of clarity, the combination of two redundant 80% reliability satellites would already provide 96% mission reliability.

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Fig. 1. Weibull distribution of satellite reliability along its lifetime: given in [13] (continuous line) and extrapolated to 100 years (dotted line).

Although aging is unavoidable, there are efforts to use prognosis analysis [41] to identify accelerated wear-out or premature mortality in the satellite elements. This allows providing either physical or functional replacement, depending on whether the satellite is still on the ground or in orbit. Failure prediction is critical in many missions, e.g. manned flights, and very often is based on extensive processing of telemetry data, which is statistically compared to datasets obtained during component tests. Thus, comprehensiveness of the telemetry records and extensive validations tests are crucial to identify critical failures before they have catastrophic consequences.

3. Strategies and available technologies to enhance survivability The space hazards already described have been a concern for satellite designers since the beginning of the space era, so a series of techniques involving not only spacecraft design but also global mission definition have been developed to overcome the existing risks and to improve satellite survivability. 3.1. Satellite orbit The lifetime requirement usually limits the lower range of the orbital altitude, where atmospheric drag may produce unacceptable decay rates. On the other hand, radiation exposure may limit equipment reliability if the orbit is unprotected from the Earth's magnetosphere, as occurs at higher altitudes, such as in GEO positions. Finally, the orbital profile definitively drives the thermal conditions of the spacecraft. Whereas GEO positions are fixed with respect to Earth, with a quite stable space weather, LEO profiles allow a rich variety of options with respect to the space environment. Although these orbits can help to simplify many spacecraft subsystems and the launching process, their efficiency from the survivability point of view needs to be reviewed. Regarding radiation, should the mission take place in the LEO region, the Earth's magnetic field protects the satellite against most of the high energy particles coming from SPE and GCR (mainly in non-polar orbits), although it will be still exposed to radiation from the particles trapped in the Earth radiation belts. Particle flux varies with orbit inclination (Fig. 2 and Fig. 3), especially the number of electrons, which is notably lower for an equatorial orbit. The number of protons also decreases for lower

5

inclination orbits, although the difference is not large and particle energy dependant. Avoiding the South Atlantic Anomaly is relevant when looking for low radiation effects. Satellites in LEO are also affected by orbit decay. In order to reach a lifespan of 100 years, initial orbit must be high enough to avoid destructive atmosphere interaction throughout the whole mission. Fig. 4 shows that with an initial orbital height over 800 km, a lifespan of 100 years is achieved with even unfavourable ballistic coefficients. Debris hazard is also very dependant from the selected orbit. The consensus of opinion at the NASA Orbital Debris Program Office appears to fall in the range of 0.5–1.0 cm as the standards for lethality [72]. So, assuming an average value of 0.75 cm, the impact probability in 100 years is around 0.03% (calculated for a cubesat of 0.01 m2 cross-section at 800 km height, Fig. 5). Better figures can be achieved if extra on-board protections are implemented. In the same way the altitude is affected during the lifetime by the dissipative drag forces, other orbital parameters may evolve from their nominal values due to other forces. Thus, eccentricity is influenced by solar pressure whereas conservative forces (geopotential) affect inclination, right ascension of the ascending node and perigee argument. Extensive studies cover long-term evolution of geostationary and navigation satellite orbits [3], as they are quite populated and the risk of collision is growing. These orbits normally present a direct coupling between orbital period and Earth rotation period. The conclusions show sensitive variations in eccentricity and inclination only when certain initial perturbations occur. This fact can be very useful for cheap end of life disposal [1]. Other interesting studies track the movement of dead satellites to develop long-term propagation models, demonstrating the stability of frozen orbits and explaining resonances in parameter fluctuations caused by lunar pole precession period [70]. For low Earth orbits with arbitrary periods, the secular variations of classical parameters are well known. Atmosphere drag tends to circularize the orbit whereas solar pressure produces noticeable eccentricity excursions only when ballistic coefficients are unfavourable. On the other hand, the oblateness of the Earth, the Sun and the Moon move the right ascension of the ascending node and the perigee with rates that are a function of the semi-major axis, inclination and eccentricity. This enables the sun-synchronous configurations and other more sophisticated frozen conditions [6], where nominal parameters are selected to nullify the main perturbation terms. Although the precise maintenance of the orbital configuration could not be extended for a long period without active means on-board, passive spacecrafts could enjoy a reasonably stable thermal environment if the orbit keeps eclipse periods and Earth distance unchanged. In this case, the most appropriate orbits from the thermal point of view are those without eccentricity and with low inclination. The circular configuration ensures minimum impact of solar pressure on eccentricity and a stationary Earth view factor. The equatorial configuration stabilizes the eclipse/sunlight pattern against seasonal effects and small changes in inclination, which make the perturbations on the ascending node and perigee irrelevant. Regarding all these aspects, a low Earth orbit seems to be a good choice for extended lifespan missions; optimum inclination should be 01 (minimizing the radiation dose) but given that it could be difficult to find a launcher to this orbit, a 281 inclination value could also be considered.

3.2. Size, shape and shielding A relevant design factor is the spacecraft size, shape and shield. These aspects have an important influence on the drag, thermal equilibrium or radiation and debris mitigation.

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

6

Fig. 2. Graphic representation of proton fluence per square centimetre for a 100-year lifespan, 800-km height circular orbit at various inclinations. South Atlantic Anomaly is conspicuous. Source: ESA SPENVIS 2012 [22].

Integral fluence (cm-2s-1)

Trapped particles (circular, 800-km height)

Protons, i=98 deg Protons, i=28 deg

1.E+16

Protons, i=0 deg 1.E+15

Electrons, i=98 deg Electrons, i=28 deg

1.E+14

Electrons, i=0 deg

Height (km) 850

Initial height for a 100-year lifespan (i=28deg)

800 750 700 650 600

1.E+13

550 500

1.E+12

0 1.E+11

0.015

0.02

Fig. 4. Orbital height needed for 100-year lifespan for different ballistic coefficients. Source: NASA DAS 2013 [44].

1.E+10 1.E+09 0.01

0.005 0.01 Ballistic coefficient (m2/kg)

0.1

1 10 Particle energy (MeV)

100

1000

Fig. 3. Trapped particle integral fluence for a 100-year lifespan, 800-km height equatorial orbit. Source: ESA SPENVIS 2012 [22].

Hardware shielding is a common practice in most spacecrafts, as it is of paramount importance in their protection against debris impacts and radiation. The spacecraft configuration provides a first mechanism for radiation protection, as more delicate pieces of equipment can occupy inner positions whereas other more resistant elements can act as natural shields without increasing the mass. In general, larger satellites can be better protected [20]

while small sizes may involve less flexibility in the configuration design. Sometimes miniature components are mandatory, which are more prone to radiation caused failures, and some others there is less room for dedicated shields. On the other hand, large satellites can suffer from unexpected radiation effects from secondary particles, due to scattering from some parts to others. The most common material for the shields is aluminium, a low-density material with excellent structural properties and very efficient when built in layers. Current efforts are focused on getting new materials (like polyethylene) with better radiation blocking capabilities [8] without showing too much degradation. Fig. 6 and Fig. 7 show that most of the protective effect can be achieved with just a 2 mm thickness shield. The remaining radiation is very

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎ Total dose (rad)

7

Total dose at 800-km height (i=0 deg, Si target, Al shield sphere)

Trapped protons Trapped electrons

1.E+06

Bremstrahlung Total mission

1.E+05 1.E+04 1.E+03 1.E+02 1.E+01 1.E+00 1.E-01 1.E-02 0

5

10

15

Shield thickness (mm)

Fig. 6. Total dose of particles at 800-km height equatorial orbit along 100 years for different shield thicknesses. Source: ESA SPENVIS 2012 [22].

Fig. 5. Number of impacts vs orbital height for different particle diameters during a 100 years mission. Source: NASA DAS 2013 [44].

difficult to block, being a mix of higher energy particles and the secondary radiation originated by the interaction between the incoming particles and the shield. Regarding physical impacts from debris and micrometeoroids, a low cross section with an aluminium shield would allow dealing with the problem in two ways: protection and avoidance. Minimizing the size (cross section) of the spacecraft is a passive collision avoidance method and a bumper shield will help to ensure spacecraft survivability against impacts of small fragments. The consequences of meteoroid and orbital debris penetrations and their link to catastrophic failure are mainly related to three factors: size, velocity and angle of incidence of the debris against the shield surface [21]. Extensive research has been done in the development of effective ways to shield manned and unmanned spacecraft against such collisions. The range of shield designs is wide, from titanium alloys [34] to honeycomb and multi-layer insulation (usually based on aluminium alloys with additional Kevlar 310 and/or Betacloth layers) [52]. 3.3. Fault tolerance: redundancy and reliability Fault tolerance protects against a wide spectrum of potential faults mainly produced by radiation and other malfunctions. The fault tolerance process usually involves several phases, i.e. error detection, reconfiguration, recovery and re-establishment of normal operations. The simplest but still efficient way to ensure survivability against malfunctions is redundancy. Most of the electronic devices on-board can be easily duplicated without a dramatic impact in the overall mass. Depending on the criticality of the element's function, cold or hot redundancy can be selected at design time. In the first case the backup hardware is switched off or put into hibernation, waiting for the on-board authority to start up and to command the new duty. In the second case, the redundant piece of equipment is working in parallel with the nominal one, executing the same tasks. It is frequent that each redundant device monitors the other's activity, being able to execute actions upon malfunction detection. During a hot redundancy operation, a cross-trap

mechanism allows the selection of the best output from the available ones. Needless to say, these selectors, normally electronic switches with some extra harness, have to present a superior reliability not to jeopardize the redundancy efforts. Whereas hot redundancy provides immunity against single failures at a price of power consumption, cold redundancy involves the need of some reaction time to set up the backup configuration on, reducing system availability. Radiation hardening is the process of protecting space components from the novice effects of radiation. The process is commonly developed in two ways: Radiation Hardening by Process [18], which takes into account the base technology to be used for the development of silicon base integrated circuits, and Radiation Hardening by Design [14], which applies dedicated design methodologies for the same purpose, for example implementing error detection and correction circuits. Most devices that are developed for space applications have been designed and produced following both approaches; it is quite easy to find space certified memory devices based on Flash and other EEPROM technologies [5], usually twined to fieldprogrammable gate arrays (FPGA) [4], with internal mechanisms to ensure fault tolerance [33]. However, it is worth saying that a new technology known as FeRAM (using ferroelectric materials) is now being seriously considered as an alternative for space applications. These are easy to use, non-volatile, low power consumption and endurance extended. Today standards for top density rounds 128 Mb per chip, as prototyped by Toshiba [60], but most of the available ones are in the 1–8 Mb range from manufacturers such as Ramtron or Fujitsu. The highest obstacle at the moment in using FeRAM devices in spacecrafts is the lack of commercially available radiation hardened components. Although their own nature makes the memory cell itself immune or highly resistant to heavy ioninduced errors, the surrounding logic built using a standard CMOS process could fail to operate in the space environment. This is something that could be solved in the short term but, at the moment, it forces designers to make use of standard hard-rad electronics already certified for space use. According to Fig. 4, the TID for a satellite in the equatorial orbit after 100 years could be equivalent to the TID for a GEO one after 15 years (15 krad), and much lower than TID received by spacecrafts conducting scientific missions far from the Earth [36]. The technologies and components already used by GEO satellites should therefore fit well in

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

8

Total dose (rad)

Total dose at 800-km height (Si target, Al shield sphere)

i=0 deg i=28 deg

1.E+07

i=98 deg

1.E+06

demanded. It is well known that solar cells suffer from degradation due to several factors: radiation, UV light, hot spots and thermal cycling and physical damage caused by debris or meteoroids. With respect to radiation, Fig. 8, obtained using methods reported in [22], shows the expected degradation in maximum power after the total mission lifespan when using GaAs/Ge materials. This material is one of the most affordable and known

1.E+05

1.E+04 0

5

10

15

Shield thickness (mm) Fig. 7. Total dose of particles at 800-km height along 100 years for different orbital inclinations and shield thickness. Source: ESA SPENVIS 2012 [22].

the proposed mission, and would enable us to avoid more specific components designed for some outer space scientific missions. As explained in [65], a way to enhance the lifetime of a component is to make it work within derating requirements, that is to say well below its normal limiting capabilities. Although derating may lead to over-dimensioned satellites, by separating the average stress levels from the strength levels of the electromechanical components, the probability of exceeding the latter reduces. The parameters where derating is normally applied include mechanical stress, temperature, power dissipation, current and voltage. The type and nature of the component lead to different recommendations for derating. Noticeably, depending on the component, there are some parameters that are not recommended for derating (e.g. voltage discharge rate in conventional capacitors). Finally, the concept rejuvenation deserves a mention within the techniques to improve the system reliability after the years. Redundant systems, either cold, warm or hot, may help each other to recover fresh status after some failures in a rejuvenation process. This is typically applied to software and processors [37] but could also be applicable to other electronic assets. Rejuvenation improves reliability at the expense of availability, which is often a very relaxed requirement in long missions. 3.4. Power sources Nuclear energy, both in the form of radioisotopes and reactors, has proven to be a lasting source of power for on-board systems. Some of the space probes launched 40 years ago are still working, even close to the edge of the solar system – if not out of it. However, current regulation constraints make it difficult to envisage nuclear-powered missions in Earth's orbit [71]. The cost is also considered unaffordable for small missions. Primary batteries could be considered to store all the required energy at launching time, so that extra charge is not needed anymore. Integrity and internal impedance of the battery should ensure the provision of required power as the mission develops. A promising option is betavoltaic cells [26], but although these devices could provide a 100-year lifespan, their power supply is not large enough to serve the whole mission. The currently available figures are around tenths of μA. Photovoltaic cells are, on the other hand, the most popular device to harvest energy in space. Solar radiation is caught and converted into electrical power, which can be used or stored as

Fig. 8. Normalized degradation in maximum power for GaAs/Ge solar cells as a function of coverglass thickness along 100 years in equatorial, 281 and polar inclinations (from top to bottom) at 800-km height. Source: ESA SPENVIS 2012 [22].

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

1 MeV equivalent electron fluences (cm-2)

GEO

1.E+16

EOL power (% from BOL)

100%

LEO polar

9

1.0% per year 1.5% per year 2.0% per year

LEO i=28 deg LEO equatorial

2.5% per year

75%

1.E+15 50%

1.E+14

25%

0%

1.E+13 10

100 1000 Coverglass thickness (µm)

10000

Fig. 9. Comparison of electron fluences affecting GaAs/Ge solar panels for several orbits as a function of coverglass thickness. Source: ESA SPENVIS 2012 [22].

technologies for photovoltaic cells. Equatorial orbit is better suited for this application, since particle fluence is much lower than that in 281 inclination, which involves crossing the South Atlantic Anomaly. In any case, the degradation is low due to the protection of Earth magnetosphere. If polar orbits are selected, bottom plot of Fig. 8 shows slightly higher degradation, which is nonetheless much friendlier than what can be expected in geo-stationary orbit, as demonstrated in Fig. 9. Solar proton flares can be dangerous, but their effect is less important in low Earth orbits. Additionally, UV light from the Sun is unavoidable for solar panels, and other cause of performance degradation. However, some new transparent silicones perform very well as coverglass for the active cells, reaching figures not higher than 3% for 15-year operation [76], that is to say 0.2% a year. A hot spot situation occurs when a solar cell within a module generates less current than the string current of the module or of the photovoltaic generator. This is typical when some cells are shaded or damaged. The shaded cell becomes reverse biased and dissipates power in the form of heat, degrading or destroying itself and its surroundings. Finally, impact with debris and meteoroids inflicts physical damage to the panels and its cells. Delamination, glass fracture, diode or inverter malfunction or disconnection are some of the observed damages. Long-lasting missions like MIR, IUE and Hubble have provided valuable statistical data [61]: after 10-year operations, surface damage was in the order of 0.045% due to objects larger than 1-mm diameter and 0.01% due to smaller ones. By extrapolating those figures to 100 years, degradation by impacts could still be bounded below quite low figures, although special attention must be paid to the electrical structure of the array to avoid cascading effects and hot spots. Regarding panel thermal environment, the same concerns as for the rest of on-board equipment exist. A low Earth orbit provides extra protection at the price of involving a daylighteclipse cycle that needs to be compensated. To summarize, by selecting an equatorial orbit below  900 km height, the fluence of trapped particles decreases and the solar panel degradation reduces to a minimum. This minimum could be about 1% per year, or 1.5% to include safety margins, that is to say a ratio end/beginning of life of about 20% (Fig. 10) for a 100-year mission. Alternative sources of energy include thermo-electric mechanisms. The heat exchange due to radiation from/to Sun, Earth and deep space produces a temperature distribution all around the

0

25

50 75 Mission lifespan (years)

100

Fig. 10. Solar cell power ratio end/begin of life for several degradation rates and mission lifespan. Source: ESA SPENVIS 2012 [22].

satellite. Engineers have demonstrated that this thermal gradient can be converted into electrical power taking advantage of the Seebeck effect, by which some materials are able to generate a voltage per unit of temperature difference (Seebeck coefficient, in V/1C). For example, the raw bismuth telluride material has  287 μV/1C for certain compositions. By stacking layers of this material, a global figure over 100 mV/1C has been reported. The efficiency is very low as the temperature gradients are not high but research in materials is improving this level of efficiency. For temperature differences of a few tenths of degrees, only few milliwatts can be obtained, and hence this technique is not considered by now. Nevertheless, it is interesting as a continuous power source and could be used, for example, to command the waking up of tan hibernating satellite. Fuel cells are becoming an area of great interest for most of the space agencies. Although known and improved for decades, today's use in ground applications makes them a candidate to reach popularity in a few years time. The process allows the conversion of hydrogen into electricity (and water) without using moving parts apart from the reactant and oxidant flows, and thus, in a very quiet and reliable manner. The efficiency is high so one could consider carrying a big tank of hydrogen to serve the whole mission, although the continuous and unavoidable leaking would be a problem. In any case, the practical implementation of the explained basic chemical principle is not obvious and technology is not mature enough to make fuel cells the primary source of energy for a satellite. Other chemical power sources would require a conversion process based on movable parts, which in principle is not compatible with the idea of sleeping for 100 years and then working for a while. Ideas may run from using hydrazine rockets to run an electrical generator by pulling a rope–pulley assembly or to heat a chamber attached to the hot side of an external combustion engine such as Stirling ones. In any case, the lack of previous experiences in the use of these devices in long-lasting missions would make the proposal unreliable. 3.5. Onboard power storage Batteries are the classical devices used to store the energy surplus from the power sources. This is mandatory when operation during eclipse is requested and solar cells are the power source. Thus, the so-called secondary batteries are forced to a charge/discharge cycle during the whole mission. The depth of

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

10

Power budget in Duty phase

W

Table 2 Performance of different materials for flywheel rotors [40].

Solar panel supply

45

Consumption

Material

Tensile strength (GPa)

Density (kg/m3)

Esp

40

E-glass S-glass Kevlar Spectra 1000 T-700 graphite Managing steel

3.5 4.8 3.8 3.0 7.0 2.7

2540 2520 1450 970 1780 8000

190 265 370 430 545 47

35 30 25 20 15

discharge during the cycle often drives the battery lifespan. Unfortunately, many missions end up due to battery death. Nowadays the preferred technologies for space batteries are Ni–H2 and Li-Ion due to its high energy density. For example, Hubble has demonstrated the suitability of Ni–H2 for long-lasting missions. Charging cycles reduce battery capacity, so, if an extremely long mission is required, the charging/discharging policy must limit the number of cycles to a minimum. Besides, the depth of discharge should not be aggressive. In essence, looking for long endurance, the battery has to be just stored most of the time with a certain amount of charge. From time to time, discharge would occur, following by a charge period in which hibernation level is again reached. Needless to say, that thermal conditions should be as stable (and low temperature) as possible to minimize component aging. Although the evolution of battery capabilities in the last years has been impressive, mainly driven by the democratization of portable equipment, the aging problem is not solved yet. Batteries are currently the most probable source of satellite failure after a 15-year operation [12]. As the batteries are one of the keystones for the system to reach 100 years operation, mechanical energy storage can be an option. Flywheels are rotatory devices that absorb electrical energy by accelerating a rotor (like a motor) and deliver the energy by decelerating it (like a generator) [40]. The accumulated energy is proportional to the rotation speed and the inertia momentum of the rotor, in its turn function of its mass and shape. The system, when using magnetic bearings in space vacuum, provides relevant efficiency and virtually no aging degradation [67]. The storage is given by the strength of the material that ends up with a maximum boundary for the revolutions. A qualitative factor Esp (W h/kg) is given in Table 2. The disadvantage of flywheels is the lack of previous experience in using flywheels in long-lasting missions; in any case, small flywheels can be used to significantly reduce the cycling of batteries in low Earth orbits, where eclipses are present in every orbit. In the present mission, modern super-capacitors could also do the task once they are more tested and certified for space applications. For the reference mission, attending to mission requirements in Table 1, during the Duty phase up to 160 10-min passes may be required to safely download the data. This is the most demanding stage in terms of power, considering that systems may be in some way degraded. Assuming a power consumption of 40 W during payload downloading and 5 W during hot standby, and given that the data is to be sent in sunlight conditions, a profile like the one in Fig. 11 can be obtained. The end-of-life capacity of solar panels would be in the order of 20% (Fig. 10), so in a small satellite with 0.6 m2 panel, the maximum power instantaneously available would be about 30 W. Batteries, flywheels or capacitors should be operative, power excess from panels could be transferred to the on-board equipment to perform the operation as predicted. In another case, the operation should be constrained to the realtime power available, in such a way that the satellite would wake up from full hibernation in each orbit, although status variables can be stored in FLASH modules for a safe activity recovery.

10 5 0 0

W

10

20

30

40

50 60 Time (min)

70

Power budget in Duty phase

80

90

100

110

Solar panel supply

45

Consumption 40 35 30 25 20 15 10 5 0 0

10

20

30

40

50 60 Time (min)

70

Power budget in Duty phase

W

80

90

100

110

Solar panel supply

35

Consumption 30 25 20 15 10 5 0 0

10

20

30

40

50 60 Time (min)

70

80

90

100

110

Fig. 11. Onboard power profiles in ‘Duty’ phase. Batteries can help to transfer energy from a generation time to a different consumption time, which can be located at the end of the charging process. The bottom image shows the power constraints imposed by a no-battery operation.

The key of an effective power budget is the consumption during hibernation or standby, since in all the phases the duty cycle is very short. A completely disconnected satellite can pose difficulties for thermal control and hence reliability. A dedicated hibernation power management shall allow for critical tasks, trying to keep into service the power storage means, which are typically the oversized batteries. Flywheels can help although their reliability is not proven yet.

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

3.6. Data communications From the telecommunication point of view, being able to exchange data with a satellite after 100 years is an issue. Ground segments become obsolete after a certain number of years, entering into retrofit programs to be adapted to new missions, including changes in baseband equipment, mechanical elements and of course protocols and formats, as they look for better performance in terms of data volume, data rate, latency and reliability [15]. Very often, it is expensive to take them back to their former configuration, as recently demonstrated by the International Sun–Earth Explorer (ISEE-3) mission from NASA, returning to Earth after 30 years in space and unable to establish contact [28]. Coding schemes and standards evolve with electronic advances, and all these changes need to be implemented in modern ground facilities, more automated and more versatile than ever. Moreover, data security breaches in the last years have raised popular concerns of privacy to a level of significance previously unseen. But the steep improvement, of paramount importance for operational stations, makes it difficult to foresee standard communication schemes that are operative after a number of decades. When mission lifetime is the priority, and it is assumed that the reception of the data is to take place in the future by means of ground facilities that today are not even envisaged, simplicity must be prioritized against performance. For example, traditional modulation schemes such the phase-shift keying (PSK) could be a good base for the development of a downlink protocol. The coding technique is critical, as the receiver needs to know how to decode in order to recover the information. The satellite should provide self-contained simple data headers, so that the receiver can start a basic dialogue to decide sizes, formats or other communication options as error correction coding. VHF or S-band could serve as accessible carriers, considering that higher bands involve more complex electronics on-board. Furthermore, recent work has developed protocols for telemetry and telecommand that ensure upper-compatibility with the basic data formats used by earlier spacecrafts [69], supported by most of the worldwide space agencies. These protocols are based on a data unit called Transfer Frames. The coding provides the additional functions necessary for their transmission, such a synchronization of symbols and frames, error correction, coding and decoding. Such initiatives provide the basis for robust protocol developments and good standards for designing communication systems being in force for a long time.

4. Discussion on long lifespan satellite concepts The reference mission has a payload consisting of a mass memory block with its management intelligence, including fault

11

tolerance mechanisms, and a communication system. Several nonvolatile banks can be implemented in parallel to help with the endurance constraints. As an example, four modules may be considered in such a way that the management electronics, also fully redundant, is in charge of checking the integrity of the data, replicating right packages on top of corrupted slots if possible or isolating them if unrecoverable. Five blocks working in parallel would ensure that even with an individual reliability reduced to 60%, the memory function is met with a reliability of 99%. EEPROM/FLASH memories should consume almost no power during hibernation, allowing the storage of status information to correctly continue activities after every wakeup. This is also compatible with the need for a good thermal stability to enlarge the lifespan. The bulk data shall be loaded before launching. However, some small pieces of information can be upgraded once in orbit, as contacts are programmed on a quarterly basis. By the end of the mission, message's downlink should be possible attending to certain rules as conspicuously detected data carriers, clear codifications, self-content metadata, ciphered and open formats and careful details on message senders and intended receivers. It is possible to envisage different spacecraft concepts matching the requirements described in previous sections. In general, the technologies involved are conventional [9], except from the extra shielding (and hence mass) needed to survive such a long lifetime and specially hardened electronics. Two of the most interesting ones are below depicted. 4.1. Flat plate shape A polygonal – or circular – shaped spacecraft (as in Fig. 12) equipped with a gravity boom is a design showing multiple advantages. This passively stabilized spacecraft eases a favourable thermal environment, since the boom side is always pointing to Earth and the top face is most of the time looking at deep space or the Sun, being convenient for the solar panels. A slow rotating along the boom could also help with thermal control, although this could require some active actuators. The disc shape minimizes the cross section, getting better ballistic coefficients and reducing the amount of impacts with space debris, which should involve homogeneous degradation if satellite is slowly rotating, improving global survivability. As explained, solar panels suffer important degradation due to radiation, UV light, hot spots and thermal cycling; some of these issues could be delayed by providing a physical barrier, although obviously that would prevent them for working. Although mechanisms and moving parts are not desirable, the release of the shutter just before the Hello phase would provide ‘fresh panels’. Different ideas arise at this point: the first option is installing a shield covering part of the panel and only leaving a small fraction exposed in order to provide the minimum

Fig. 12. Artist view of flat-boomed satellite with deployable panels for end-of-mission operation. Source: Aeroxpace Worldwide Inc.

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

12

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

to reach the desired reliability, although solar panels will be permanently exposed to environment in such a way that power supply at the end of the mission can be jeopardized if panels are not duly oversized. The thermal environment can be quite stable and homogeneous, whereas compactness can help radiation protection for more sensitive components, which should occupy the centre of the structure. For similar masses, the larger cross section of this concept will lead to worse ballistic coefficients, which could make the spacecraft more sensitive to orbit decay. Maximum probability of success can be achieved by using a twin-spacecraft mission. Two redundant satellites provide extra system reliability. This approach will not duplicate global mission costs, as a very important part of them are related with spacecraft design, facilities, etc. which are non-recurrent expenses. This approach would help to overcome most of the concerns related to single events like strong debris impacts or hardware malfunctions.

5. Conclusions

Fig. 13. Artist view of tetra-decahedron satellite. Source: Nanosat 1b from INTA.

power required during the Trip phase. Other more elaborated options involve the inclusion of a deployable structure which, when folded, acts as protection and then, when deployed, provides maximum area and new panels. The most serious concern in both approaches is on the reliability of the deployment mechanism after such a long period of time. However, the simplicity of the process and the implementation of multiple redundancies should help to overcome this issue. Regarding the thermal control, a very low temperature range and cycling will help with lifespan extension in every component, especially those that are part of the power subsystem. Having selected a low orbit and maintaining a constant satellite attitude with respect to the Earth, thermal equilibrium around ambient temperature should not be difficult to achieve passively. This concept develops its full potential when locating the spacecraft in an equatorial orbit, although it is also valid for low inclination orbits, where the loss of effective area exposed to the Sun is small. The extreme lifespan considered in this paper, with more than 500,000 sun/eclipse cycles, require a thermal control that limits the internal temperature changes to a minimum, preferable using passive means. Preliminary simulations for a 10-kg flat shaped satellite in equatorial LEO, low absorptivity surface towards Earth and solar panels on the back, shows variations in the order of þ/7 3 K over a 300 K average each orbit. This is low enough to ensure stability over a long lifetime. If changing to 281 inclination, results are very similar along the whole mission. With higher inclinations, the right ascension of the ascending node drifts substantially along the mission, so although the orbital thermal cycles are shallow, the equilibrium temperature varies a lot with the months, becoming less advisable. 4.2. Tetra-decahedron shape A more conventional concept is the development of a monolithic spacecraft with the solar panels located on all its faces (Fig. 13). Again, attitude stabilization would be passive thanks to a boom. In normal cases, rotation axis is not the Sun line and the exposure of panels to the light is homogeneous, as well as the degradation due to structural damage. In comparison to the former proposal, this concept prioritizes simplicity. The lack of actuators or deployable components helps

The aggressive space environment is a major constraint for satellite lifespan. Recent missions implement mechanisms to enhance endurance by means of techniques to protect themselves from radiation effects, debris and micrometeoroid impacts, atmospheric drag, thermal fluctuations and other accelerated aging sources. A full review of hazards reveals that the challenge of setting up a mission with an autonomous lifespan of a full century is close to being achievable. As a guidance thread for the review, the preliminary feasibility of a 100-year lifespan mission is studied. Radiation, in all its forms, can be limited by using non-polar low Earth orbits, so that planet magnetosphere acts as a natural shield. Satellite architecture is important to protect more sensitive parts inside, exposing to the exterior the ones that are most resistant. Ad-hoc shields can be installed, built from aluminium or other synthetic materials. Electromechanical components, working under derating requirements, need in any case to be hardened and validated to allow for prognosis analysis during operation. Reliability enhancement and fault tolerance techniques are common practices at component, equipment and system levels. Debris and micrometeoroid impacts jeopardize the satellite at every time. Although there is a current trend to introduce mission disposal manoeuvres, the fact is that there are few interesting orbital profiles where debris is not a major concern, especially when an extended endurance is pursued. Unconventional orbits like low equatorial ones are less populated and hence improve the survivability. Physical protection may be also used to stand for more impact without degrading or destroying on-board equipment. There are good models to predict satellite drag coefficients. Unfortunately, the uncertainty of atmosphere composition, density and movements makes it difficult to accurately predict decay rates for long missions. In any case, the shape and mass of the satellite may improve the ballistic coefficient which, together with a high enough orbit, ensures long orbiting periods. With respect to satellite subsystems, the power modules, normally composed of both solar panels and batteries, are the weakest from an endurance point of view. The review of the state of the art has shown that, even though panels could work under sensitively degraded conditions after a long time, current batteries are not prepared for an operational life of many decades. Alternative solutions have been analysed, from flywheels and other power storages to no-battery just-in-time power consumption. Should a century lifespan mission be envisaged, an equatorial low Earth orbit is proposed. The satellite would need to be passively stabilized, and with a shape with a low cross-section

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

area to minimize atmospheric drag and debris impact probability. Dedicated shields would protect against trapped radiation and internal redundancy and rejuvenation mechanisms would allow lifespan extension by scheduling long hibernation periods. Passive selective surfaces provide low depth thermal cycling which minimizes material and electronic performance degradation. Power could be obtained from stacked solar panels, which could be progressively unfolded, whereas the consumption should be adapted to availability in real time. This review work demonstrates the capability of current technologies to develop unprecedented robust and autonomous space systems.

Acknowledgements The authors would like to express their deepest gratitude to Aeroxpace Worldwide Inc. for the opportunity given to be part of the amazing Space Mailbox initiative. References [1] Alessi EM, Rossi A, Valsecchi GB, Anselmo L, Pardini C, Colombo C, et al. Effectiveness of GNSS disposal strategies. Acta Astronaut 2014;99(1):292–302. [2] Anderson PV, Schaub H. Local debris congestion in the geosynchronous environment with population augmentation. Acta Astronaut 2014;94 (2):619–28. [3] Anselmo L, Pardini C. Long-term dynamical evolution of high area-to-mass ratio debris released into high Earth orbits. Acta Astronaut 2010;67:204–16. [4] Arbat A, Calligaro C, Dayan V, Pikhay E, Roizin Y, Non-volatile memory for FPGA booting in space. In: Proceedings of the 2013 NASA/ESA conference on adaptive hardware and systems, AHS 2013, Article number 6604247, 2013, p. 204–8. [5] Arbat AA, Calligaro CA, Dayan VB, Pikhay EB, Roizin Y. SkyFlash EC project: architecture for a 1 Mbit S-Flash for space applications, 2012. In: 19th IEEE international conference on electronics, circuits, and systems, ICECS 2012, Article number 6463671, p. 617–20. [6] Aorpimai M, Palmer PL. Analysis of frozen conditions and optimal frozen orbit injection. J Guid Control Dyn 2003;26(5):786–93. [7] Bird GA. Molecular gas dynamics and the direct simulation of gas flow. Oxford: Clarendon Press; 1994; 42. [8] Borggräfe A, Quatmann M, Nö lke D. Radiation protective structures on the base of a case study for a manned Mars mission. Acta Astronaut0094-5765 2009;65(9–10):1292–305. [9] Bouwmeester J, Guo J. Survey of worldwide pico- and nanosatellite missions, distributions and subsystem technology. Acta Astronaut 2010;67(7–8): 854–62. [10] Bowman BR, Tobiska WK, Marcos FA, Valladares. The JB2006 empirical thermospheric density model. J Atmos Solar-Terr Phys 2008;70(5):774–93, http://dx.doi.org/10.1016/j.jastp.2007.10.002. [11] Burghardt C, Meissner F, Enke H, Oehler S, Letschnik J. Propellant endurance prognosis for GEO operations. In: Proceedings of the SpaceOps 2012 Conference; 2012. [12] Castet JF, Saleh JH. Satellite reliability: statistical data analysis and modeling. J Spacecr Rocket 2009;46(5):1065–76, http://dx.doi.org/10.2514/1.42243. [13] Castet JF, Saleh JH. Single versus mixture Weibull distributions for nonparametric satellite reliability. Reliab Eng Syst Saf 2010;95(3):295–300 (march). [14] Chen J, Chen S, Liang B, Liu B, Liu F. Radiation hardened by design techniques to reduce single event transient pulse width based on the physical mechanism. Microelectron Reliab0026-2714 2012;52(6):1227–32. [15] Cheng MK, Lee DK, Estabrook P, Shaw H, Wong Y, Flynn M. An evaluation of new coding and modulation schemes for NASA Space Network. In: Proceedings of the 12th international conference on space operations; 2012. [16] Committee on Implementation of a Sustainable Land Imaging Program, Landsat and Beyond: Sustaining and Enhancing the Nation's Land Imaging Program, isbn:987-0-309-29001-2; 2013. [17] Davis DH. Monte Carlo calculation of molecular flow rates through a cylindrical elbow and pipes of other shapes. J Appl Phys 1960;31(7):1169–76. [18] Derbenwick GF, Gregory BL. Process optimization of radiation-hardened CMOS integrated circuits. IEEE Trans Nucl Sci 1975;22(6):2151–6, http://dx.doi.org/ 10.1109/TNS.1975.4328096. [19] Dever J, Banks B, de Groh K, Miller S. Degradation of spacecraft materials. In: Kutz Myer, editor. Handbook of environmental degradation of materials. Norwich, NY. ISBN: 9780815515005. p. 465–501 [chapter 23]. [20] Dubos GF, Castet JF, Saleh JH. Statistical reliability analysis of satellites by mass category: does spacecraft size matter? Acta Astronaut 2010;67:584–95. [21] Christiansen EL. Meteoroid/debris shielding. Houston, Texas: NASA Johnson Space Center; 2003 (NASA TP-2003-210788). [22] ESA SPENVIS space environment information system, release 4.6.6; April 2012.

13

[23] European Space Agency, ESA Media Relations Office. (2014). ESA's ‘sleeping beauty’ wakes up from deep space hibernation [Press release]. Retrieved from 〈http://www.esa.int/Our_Activities/Space_Science/Rosetta/ESA_s_sleeping_ beauty_wakes_up_from_deep_space_hibernation〉 on January 2014. [24] Ferri P, Accomazzo A, Hubault A, Lodiot S, Pellon-Bailon JL, Porta R. Rosetta enters hibernation. Acta Astronaut 2012;79:124–30. [25] Finkleman D. The dilemma of space debris. Am Sci 2014;102(1):26–33. [26] Gao H, Luo S, Zhang H, Wang H, Fu Z. Demonstration, radiation tolerance and design on a betavoltaic micropower. Energy0360-5442 2013;51:116–22. [27] Gaposchkin EM, Coster AJ. Analysis of satellite drag. Linc Lab J 1988;1:203–24. [28] Garcia L, Farquhar R, Eastman T. New opportunities for a historic spacecraft. Space Weather 2012;10(8). [29] Klinkrad Heiner. Space debris: models and risk analysis. Berlin, Springer; 2006. [30] Hill ME, Decker RB, Brown LE, Drake JF, Hamilton DC, Krimigis SM, et al. Dependence of energetic ion and electron intensities on proximity to the magnetically sectored heliosheath: Voyager 1 and 2 observations. Astrophys J 2014;781(2) (art. no. 94). [31] Inter-Agency Space Debris Coordination Committee (IADC). Stability of the future LEO environment, IADC-12-08, Rev. 1 January 2013. [32] Inter-Agency Space Debris Coordination Committee (IADC). Support to the IADC space debris mitigation guidelines, IADC-04-06, October 2004. [33] Jacobs A, Cieslewski G, George AD, Gordon-Ross A, Lam H. Reconfigurable fault tolerance: a comprehensive framework for reliable and adaptive FPGA-based space computing. ACM Trans Reconfig Technol Syst 2012;5(4):1–30. [34] Iyer Kaushik A, Poormon Kevin L, Deacon Ryan M, Mehoke Douglas S, Swaminathan PK, Brown Robert C. Hypervelocity impact response of Ti–6Al–4V and commercially pure titanium. Procedia Eng 2013;58:127–37. [35] Kelly FP, Holm TM. Landsat: sustaining earth observations beyond Landsat 8. Photogramm Eng Remote Sens 2014;80(1):15. [36] Keys AS, Adams JH, Darty RC, et al. Exploration technology development program's radiation hardened electronics for space environments (RHESE) project overview. International planetary probes workshop – 6 Atlanta, GA, USA; 2008. [37] Kim DS, Lee SM, Jung J-H, Kim TH, Lee S, Park JS. Reliability and availability analysis for an on board computer in a satellite system using standby redundancy and rejuvenation. J Mech Sci Technol 2012;26(7):2059–63. [38] Lakshminarayana V, Karthikeyan B, Hariharan VK, Ghatpande ND, Danabalan TL. Impact of space weather on spacecraft. In: Proceedings of the international conference on electromagnetic interference and compatibility; 2008, art no. 5154337, p. 481–6. [39] Laporte F, Sasot E. Operational management of collision risks for LEO satellites at CNES. In: Proceedings of the SpaceOps 2008 conference; 2008. [40] Liu H, Jiang J. Flywheel energy storage – an upswing technology for energy sustainability. Energy Build 2006:599–604. [41] Losik L. Moving to the 100% reliability domain using prognostic analysis. In: AIAA SPACE Conference and Exposition 2010. Anaheim, CA; United States, art. code 98237; 2010. [42] Moe K, Moe MM, Wallace SD. Improved satellite drag coefficient calculations from orbital measurements of energy accommodation. J Spacecr Rockets 1998;35(3):266–72, http://dx.doi.org/10.2514/2.3350. [43] Mostaza D, Graziano BP, Roberts PCE. Spacecraft drag modelling. Prog Aerosp Sci 2014;60:56–65. [44] NASA DAS Debris Assessment Software, release 2.0.2; 2013. [45] NASA. The threat of orbital debris and protecting NASA space assets from satellite collisions; April 2009 〈http://images.spaceref.com/news/2009/ODMe diaBriefing28Apr09-1.pdf〉 [taken 20.02.14]. [46] Park SY, Choi HS, Choi WJ, Kwon H. Effect of vacuum thermal cyclic exposures on unidirectional carbon fibre/epoxy composites for low earth orbit space applications. Compos Part B: Eng 2012;43(2):726–38. [47] Peret L, Legendre P, Delavault S, Martin T., Detection of orbital debris collision risrks for the automated transfer vehicle, Proceedings of the 20th international symposium on space flight dynamics; 2009. [48] Picone JM, Heidin AE, Drob DP, Aikin AC. NRLMSISE-00 empirical model of the atmosphere: statistical comparisons and scientific issues. J Geophys Res 2002;107(A12):1468, http://dx.doi.org/10.1029/2002JA009430. [49] Regan FJ, Anandakrishnan Satya M. Dynamics of atmospheric re-entry. American Institute of Aeronautics and Astronautics. 2nd ed.. Washington, DC: SPEC-Books; 1993. [50] Remaury S, Nabarra P, Bellouard E, d'Escrivan S. In-flight aging of th1ermal coatings: THERME Experiment. J Spacecr Rockets 2011;48(1):27–33. [51] Remaury S, Guillaumon JC, Nabarra P. Behaviour of thermal control coatings under atomic oxygen and ultraviolet exposures. In: Kleiman JI, Iskanderova Z, editors. Proceedings of the 6th international conference on protection of materials and structure from the space environment. Norwell, MA: Kluwer Academic Publishers; 2003. p. 193–201. [52] Turner Robert J, Taylor Emma A, Anthony J, McDonnell M, Stokes Hedley, Marriott Peter, et al. Cost effective honeycomb and multi-layer insulation debris shields for unmanned spacecraft. Int J Impact Eng 2001;26(1–10): 785–96. [53] Sajjadi-Kia H, Ababei C. A new reliability evaluation methodology with application to lifetime oriented circuit design. IEEE Trans Device Mater Reliab 2013;13(1):192–202 (Article number 6357233). [54] Saleh JH, Hassan R, Torres-Padilla JP, Hastings DE, Newman DJ. Impact of subsystem reliability on satellite revenue generation and present value. J Spacecr Rockets 2005;42(6):1122–9.

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i

14

J. Gonzalo et al. / Progress in Aerospace Sciences ∎ (∎∎∎∎) ∎∎∎–∎∎∎

[55] Saleh JH, Torres-Padilla JP, Hastings DE, Newman DJ. To reduce or to extend a spacecraft design lifetime. J Spacecr Rockets 2006;43(1):207–17. [56] Saleh JH, Hastings DE, Newman DJ. Spacecraft design lifetime. J Spacecr Rockets 2002;39(2):244–57. [57] Satellite thermal control handbook, 2nd ed. Gilmore D, editor. The Aerospace Corporation Press, El Segundo, California; 2002. [58] Schaaf SA, Chambre PL. Flow of rarefied gases, vol. 8. Princeton, NJ: Princeton University Press; 1961. [59] Schamberg RA. A new analytic representation of surface interaction with hypothermal free molecule flow with application to neutral-particle drag estimates of satellites, technical report RM-2313, RAND research memorandum; 1959. [60] Shiga H, Takashima D, Shiratake S, Hoya K, Miyakawa, et al. A 1.6 GB/s DDR2 128 MB chain FeRAM with scalable octal bitline and sensing schemes. IEEE J Solid-State Circuits 2010;45(1):142–52. [61] Smirnov VM, Semenov AS, Sokolov VG, Konoshenko VP, Kovalyov II. Study of micrometeoroid and orbital debris effects on the solar panels retrieved from the space station MIR. Space Debris 2000;2(1):1–7. [62] Space engineering, calculation of radiation and its effects and margin policy handbook, ECSS Std., ECSS-E-HB-10-12A, ESA-ESTEC Requirements & Standards Division, Noordwijk, The Netherlands; 2010. [63] Space engineering, methods for the calculation of radiation received and its effects, and a policy for design margins, ECSS Std., ECSS‐E‐ST‐10‐12C, ESAESTEC Requirements & Standards Division, Noordwijk, The Netherlands; 2008. [64] Space engineering, space environment, ECSS Std., ECSS‐E‐ST‐10‐04C, ESAESTEC Requirements & Standards Division, Noordwijk, The Netherlands; 2008. [65] Space product assurance, Derating – EEE components, ECSS Std., ECSS‐Q‐ST‐ 30‐11C Rev 1, ESA-ESTEC Requirements & Standards Division, Noordwijk, The Netherlands; 2011.

[66] Tafazoli M. A study of on-orbit spacecraft failures. Acta Astronaut 2008;64 (2009):195–205. [67] Tang J, Fang J, Sam S. Roles of superconducting magnetic bearings and active magnetic bearings in attitude control and energy storage flywheel. Phys C: Superconduct 2012;483:178–85. [68] The Earth Observing One (EO-1) satellite mission: Over a decade in space, Middleton EM, Ungar SG, Mandl DJ, Ong L, Frye SW, et al. IEEE J Sel Top Appl Earth Obs Remote Sens 2013;6(2):243–56 (art. no. 6510002). [69] The Consultative Committee for Space Data Systems. Space data link protocols – summary of concept and rationale. Informational report CCSDS 130.2-G-2. Green book; November 2012. [70] Ulivieri C, Circi C, Ortore E, Bunkheila F, Todino F. Frozen orbital plane solutions for satellites in nearly circular orbit. J Guid Control Dyn 2013;36 (4):935–45. [71] United Nations Office for Outer Space Affairs. Resolution 47/68 of 14 December 1992, The principles relevant to the use of nuclear power sources in outer space. [72] Vance L, Mense A. Value analysis for orbital debris removal. Adv Space Res0273-1177 2013;52(4):685–95. [73] White AE, Lewis HG. The many futures of active debris removal. Acta Astronaut 2014;95(1):189–97. [74] Woodraska DL, Woods TN, Eparvier FG. Satellite drag compared with the solar extreme-ultraviolet experiment measurements. J Spacecr Rockets 2007;44 (6):1204–9, http://dx.doi.org/10.2514/1.28639. [75] Zhang J, Hai Z, Thirugnanasambandam S, Evans JL, Bozack MJ, Zhang Y, et al. Thermal aging effects on the thermal cycling reliability of lead-free fine pitch packages. IEEE Trans Compon Packag Manuf Technol 2013;3(8):1348–57 (Article number 6490099). [76] Zimmermanna CG. Time dependent degradation of photovoltaic modules by ultraviolet light. Appl Phys Lett 2008;92:241110.

Please cite this article as: Gonzalo J, et al. On the challenge of a century lifespan satellite. Progress in Aerospace Sciences (2014), http: //dx.doi.org/10.1016/j.paerosci.2014.05.001i