Shape Memory Activated Self-Deployable Solar Sails: Small-Scale Prototypes Manufacturing and Planarity Analysis by 3D Laser Scanner

Home , IKAROS, LightSail, Sunjammer (spacecraft)

actuators

Article Shape Memory Activated Self-Deployable Solar Sails: Small-Scale Prototypes Manufacturing and Planarity Analysis by 3D Laser Scanner

Alberto Boschetto 1, Luana Bottini 1, Girolamo Costanza 2,* and Maria Elisa Tata 3

1 Mechanical and Aerospace Engineering Department, University of Rome “La Sapienza”, via Eudossiana 18, 00184 Rome, Italy; [email protected] (A.B.); [email protected] (L.B.) 2 Industrial Engineering Department, University of Rome “Tor Vergata”, Via del Politecnico 1, 00133 Rome, Italy 3 Civil Engineering and Computer Science Department, University of Rome “Tor Vergata”, Via del Politecnico 1, 00133 Rome, Italy; [email protected] * Correspondence: [email protected]

Received: 13 April 2019; Accepted: 30 April 2019; Published: 3 May 2019

Abstract: Solar sails are propellantless systems where the propulsive force is given by the momentum exchange of reflecting photons. Thanks to the use of shape memory alloys for the self-actuation of the system, complexity of the structure itself has decreased and so has the weight of the whole structure. Four self-deploying systems based on the NiTi shape memory wires have been designed and manufactured in different configurations (wires disposal and folding number). The deployed solar sails surfaces have been acquired by a Nextengine 3D Laser Scanner based on the Multistripe Triangulation. 3D maps have been pre-processed through Geomagic Studio and then elaborated in the Wolfram Mathematica environment. The planarity degree has been evaluated as level curves from the regression plane highlighting marked differences between the four configurations and locating the vertices as the most critical zones. These results are useful in the optimization of the best folding solution both in the weight/surface reduction and in the planarity degree of the solar sail.

Keywords: shape memory alloy; solar sail; nitinol; laser scanner

1. Introduction The solar sails are propulsion systems based on the radiation pressure of sunlight applied on the sail itself. At the beginning of the 1920’s, the Soviet Konstantin Tsiolkovsky and his colleague Fridrickh Tsander, who are considered the true pioneers in the astronautical field, wrote “For flight in interplanetary space, I am working on the idea of flying, using tremendous mirrors of very thin sheets, capable of achieving favorable results, using the pressure of sunlight to attain cosmic velocities” [1,2]. In 1973, NASA successfully applied solar sail during the mission of the spacecraft Mariner 10, which is dedicated to the exploration of Venus and Mercury. Until 2010, no solar sails were successfully used in space as primary propulsion systems. On 21 May, 2010, the Japan Aerospace Exploration Agency (JAXA) launched the IKAROS (Interplanetary Kite-craft Accelerated by Radiation Of the Sun) spacecraft, which deployed a 200 m2 polyimide experimental solar sail. IKAROS had a diagonal spinning square sail made of a 7.5 µm thick sheet of polyamide. The polyamide sheet had a mass of about 10 grams per square meter. A thin-film solar array was embedded to the sail. Eight LCD panels were embedded in the sail, whose reflectance can be adjusted for attitude control [3,4]. IKAROS spent six months traveling to Venus, and then began a journey towards the sun [5,6]. Despite the losses of Cosmos 1 and NanoSail-D (which were due to

Actuators 2019, 8, 38; doi:10.3390/act8020038 www.mdpi.com/journal/actuators Actuators 2019, 8, 38 2 of 10 failure of their launchers), scientists and engineers around the world have been working on solar sails while most applications created intend to use the sails as cargo transport. Due to the small pressure level, in order to exploit this propulsion system, a great surface of the sail is required. For the same reason, materials employed for the construction of the sail must show low weight and high reflectivity [7]. Engines based on ion propulsion for small satellites are able to work according to the principle that a small but constant pressure applied on a wide surface of the sail can produce a satisfying acceleration to the whole structure [8]. The planned test flight of the Solar Sail Demonstrator, called the “Sunjammer” by its designers in honor of the 1964 Arthur C. Clarke story, in which he coined the term “solar sailing”, had the intent to prove the feasibility of the sailing technology. The demonstration was expected to launch on a Falcon 9 in January 2015. It would have been a secondary payload, released after the placement of the DSCOVR climate satellite at the L1 point. Citing a lack of confidence in its contractor’s ability to deliver, the mission was cancelled in October 2014. In 2016, the Planetary Society’s LightSail 1 spacecraft launched into orbit and transmitted its first signals back to Earth, which kicked off a three-and-a-half week mission to test the CubeSat’s critical functions and deploy a 32 m2 solar sail. LightSail-1 consisted of three CubeSat units. One unit was occupied by the core, containing all the devices, the equipment, and the control systems, while the remaining two units were occupied by the four folded sail segments. A peculiarity of the project was that the CubeSat would be inserted into a small satellite, Prox 1, and together they will be launched into orbit. The goal of Prox 1 is to demonstrate the possibility of using small satellites to inspect and control other spacecrafts. Now, as engineers prepare for a full systems test of LightSail 2, a successor CubeSat will attempt the first controlled solar sail flight in low-Earth orbit. The Near-Earth Asteroid Scout (NEA Scout) has been a mission jointly developed by NASA’s Marshall Space Flight Center (MSFC) and the Jet Propulsion Laboratory (JPL), which consists of a controllable low-cost CubeSat solar sail spacecraft capable of encountering near-Earth asteroids (NEA). In 2018, NASA’s Near-Earth Asteroid Scout, which is a small satellite designed to study asteroids close to Earth, performed a successful deployment test on June 28th of the solar sail to be launched on Exploration Mission-1 (EM-1). The test was performed in an indoor clean room at the NeXolve facility in Huntsville, Alabama. In its mission, the NEA Scout will perform reconnaissance of an asteroid using a CubeSat and solar sail propulsion, which offers navigation agility during cruise for approaching the target. Propelled by sunlight, the NEA Scout will complete a flyby and observe a small asteroid (<300 feet in diameter), while also taking pictures and observing its position in the space, the asteroid’s shape, rotational properties, spectral class, local dust and debris field, regional morphology, and regolith properties. The data collected will enhance the current understanding of asteroid environments and will yield key information for future human asteroid explorers. Many different systems have been previously considered for the sails opening. Each system was characterized by the presence of guide rollers, electromechanical actuation devices, or composite booms [9]. In the actual deployment technology, the main limit is the high weight of the system and the complexity of the deployment mechanism for such huge surfaces. Objective of the present work has been to propose, manufacture, and test an innovative miniaturized self-deploying system actuated by shape memory alloy elements. A self-deploying system based on the NiTi Shape Memory wires has been designed and manufactured in a small-scale prototype. Kapton has always been employed as a sail surface with a thin Al coating. In our experiments, commercial pure Al thin sheets in addition to adhesive kapton film have been used in order to simulate the sail. By testing different configurations, an attempt has been made to improve the achievable flatness of solar sail prototypes after deployment, by varying the number and position of the shape memory elements. The prototypes made have been successful in unfolding in a single heating step, which increases the number of deployments. Through the techniques of Reverse Engineering, it has been possible to obtain virtual models of the manufactured and tested prototypes. On these models, an analysis has been performed to obtain a quantitative measurement of the degree of planarity of the prototypes, with which it has been possible to compare the performance offered by the four configurations. To compare the different configurations Actuators 2019, 8, 38 3 of 10 of the Box-and-Whisker diagrams, histograms and diagrams with level curves were used. Useful results for the optimization of the best folding solution in terms of the surface/weight reduction and in the planarity degree of the solar sail were achieved.

2. Materials and Methods Actuators 2019, 8, x FOR PEER REVIEW 3 of 10 The most frequently used materials for the manufacturing of the solar sails are Kapton® and compare the different configurations of the Box-and-Whisker diagrams, histograms and diagrams ® Mylar . Bothwith oflevel them curves belong were used. to a classUseful of results plastic for materialsthe optimization with of high the best performance folding solution in terms in terms of lightness, flexibility,of high the resistancesurface/weight temperatures, reduction and andin the chemical planarity degree agents. of the These solar materials sail were achieved. are used as substrate and then coated with a thin aluminum film to obtain a more reflective surface. The choice of the materials 2. Materials and Methods is fundamental in order to achieve the best performance in terms of a number of possible folds and ® lightness of theThe structure most frequently [10] and used active materials surface for the [11 manufacturing]. Because of of the the costs solar ofsails these are materials,Kapton and it has been Mylar®. Both of them belong to a class of plastic materials with high performance in terms of necessarylightness, to use an flexibility, alternative high resistance material temperatures for the construction, and chemical of agents. prototypes. These materials For this are reason, used as thin films (2.5 µm) ofsubstrate adhesive and kapton then coated (1.4 with g/cm a 3thin) were aluminum applied film on to obtain thin commercial a more reflective aluminum surface. The films choice (12 µm). On the aluminumof the film,materials some is fundamental active shape in memoryorder to achiev alloye the elements best performance have been in terms inserted of a number perpendicularly of to the bendingpossible line. folds As activeand lightness materials, of the structure shape memory[10] and active alloy surface wires [11]. of Because 0.41 and of the 0.6 costs mm of diameterthese have materials, it has been necessary to use an alternative material for the construction of prototypes. For been selected.this reason, Shape thin memory films (2.5 alloys μm) areof adhesive a class ofkapton functional (1.4 g/cm materials3) were applied able toon recoverthin commercial the preset shape upon heatingaluminum above films acritical (12 μm). transformationOn the aluminum temperaturefilm, some active [12 shape,13]. memory The shape alloy elements recovery have is based on the thermo-elasticbeen inserted martensitic perpendicularly transformation to the bending occurring line. As active in materials, these kind shape of memory alloys andalloy thewires characteristic of transformation0.41 and temperatures 0.6 mm diameter are have a function been selected. of the Shape composition memory alloys of the are alloy a class and of offunctional the thermal and materials able to recover the preset shape upon heating above a critical transformation temperature mechanical[12,13]. history The ofshape the recovery material is [based14]. Typicalon the thermo-elastic transformation martensitic temperatures transformation for NiTioccurring are 65in ◦C (alloy H) and 95these◦C (alloy kind of M), alloys according and the tocharacteristic the nomenclature transformation used temperatures by the supplier are a function of the alloysof the (Memory Metalle GmbH).composition A sketch of the alloy of an and aluminum of the thermal sail isand reported mechanical in history Figure of1. the It ismaterial evident [14]. that Typical the presence of siliconetransformation employed fasten temperatures the shape for NiTi memory are 65 wires°C (alloy to H) the and aluminum 95 °C (alloy sail. M), Inaccording order toto setthe the shape nomenclature used by the supplier of the alloys (Memory Metalle GmbH). A sketch of an aluminum to be recoveredsail is reported while heating,in Figure a1. particularIt is evident thermalthat the presence treatment of silicone has been employed performed, fasten the which shape is usually called shape-setting.memory wires It to consists the aluminum in heating sail. In uporder to to 500 set◦ theC while shape theto be wire recovered is kept while straight heating, on a a flat bar, maintainedparticular at this thermal temperature treatment for has 5 minutes,been performed, and quenchedwhich is usually in cold called water. shape-setting. After thisIt consists treatment, in when bent in a coldheating condition, up to 500 °C the while wire the is wire able is kept to recover straight on the a flat preset bar, maintained straight shape at this temperature upon heating for 5 above the minutes, and quenched in cold water. After this treatment, when bent in a cold condition, the wire is activationable temperature to recover the (95 preset◦C). straight shape upon heating above the activation temperature (95 °C).

Figure 1. AFigure sketch 1. ofA ansketch aluminum of an aluminum sail in conﬁgurationsail in configuration 1 (a), 21 ((ba),), and2 (b), 3 and (c). Prototype3 (c). Prototype of conﬁguration of 3 is reportedconfiguration in section 3 is (d reported). in section (d).

Three possibleThree possible configurations configurations have have been been studied studied in in which which a different a different distribution distribution of SMA wires of SMA wires and/or a different folding method are taken into account. Configuration 1 provides three and/or a dideployments,fferent folding which method take place are in takenan orthogonal into account. direction Configuration to the dotted lines 1 visible provides in Figure three 1a. deployments, In which take place in an orthogonal direction to the dotted lines visible in Figure1a. In order to make the outlines, a total of 9 NiTi wires have been used, including 4 wires with a diameter of 0.41 mm and length of 2 cm, shown in yellow, 2 wires with a diameter of 0.41 mm and a length of 4 cm, shown in Actuators 2019, 8, 38 4 of 10

green, andActuators 3 wires 2019 with, 8, x FOR a diameter PEER REVIEW of 0.6 mm and a length of 6 cm, shown in red. Following4 of 10 the 3 folds, the sail takes the form of a rectangle of dimensions 8 cm x 6 cm. The surface reduction obtained is, therefore, equalorder to to make 52%. the outlines, a total of 9 NiTi wires have been used, including 4 wires with a diameter of 0.41 mm and length of 2 cm, shown in yellow, 2 wires with a diameter of 0.41 mm and a length of In Configuration4 cm, shown in 2 (Figuregreen, and1b), 3 wires to obtain with a diameter greater percentageof 0.6 mm and of a surfacelength of reduction, 6 cm, shown the in arrangementred. of the wiresFollowing has been the 3 changedfolds, the sail in takes order the to form have of a 5 rectangle deployments of dimensions instead 8 cm of x 6 3 cm. using The surface 11 wires in the NiTi alloy.reduction Each of obtained them hadis, therefore, a diameter equal to of 52%. 0.41 mm and a length of 20 mm. The final shape of the prototype wasIn an Configuration isosceles triangle 2 (Figure with 1b), a to base obtain of 10a cmgreater and percentage height of of 5 cmsurface for areduction, surface the reduction of arrangement of the wires has been changed in order to have 5 deployments instead of 3 using 11 75%. In subsequentwires in the NiTi tests, alloy. it was Each decided of them had to replacea diameter a of part 0.41 of mm the and wires a length used of 20 in mm. configuration The final 2 with others of greatershape of length the prototype and diameter was an isosceles with thetriangle aim with toobtain a base of a 10 greater cm and degree height of of 5 planarity.cm for a surface Five wires of the previousreduction configuration of 75%. In have subsequent been replacedtests, it was by decided 4 wires to withreplace a diametera part of the of 0.41wires mm used and in a length equal to 4configuration cm, and one 2 with with others a diameter of greater of length 0.6 mmand di andameter a length with the equal aim to toobtain 6 cm. a greater In Configuration degree of 3, 11 planarity. Five wires of the previous configuration have been replaced by 4 wires with a diameter of NiTi wires0.41 have mm been and a employed:length equal to 6 4 wirescm, and with one with a diameter a diameter of of 0.410.6 mm mm and and a length a length equal to of 6 2cm. cm, shown in yellow,In 4 wiresConfiguration with a3, diameter 11 NiTi wires of 0.41have been mm employed: and a length 6 wires of with 4 cm, a diameter shown of in 0.41 green, mm and 1 wire a with a dimeter oflength 0.6 mm of 2 andcm, shown 6 cm in yellow, length, 4 wires shown with in a red.diameter This of configuration0.41 mm and a length includes of 4 cm, 5 deploymentsshown in and the surfacegreen, reduction 1 wire with is 75%,a dimeter which of 0.6 is mm the and same 6 cm asin length, for the shown previous in red. This configuration. configuration includes In addition, the 5 deployments and the surface reduction is 75%, which is the same as for the previous configuration. folding procedureIn addition, is the performed folding procedure the same is performed as in Configuration the same as in Configuration 2). In Configuration 2). In Configuration 4), the arrangement 4), of wires remainsthe arrangement the same of aswires Configuration remains the same c), whileas Configuration the folding c), while procedure the folding has procedure been changed. has In the first foldingbeen method, changed. the In lowerthe first vertices folding ofmethod, the triangle the lower have vertices been of foldedthe triangle outwards, have been unlike folded that which occurred inoutwards, Configurations unlike that 2) which and 3). occurred The second in Configurations and third step2) and remained 3). The second unchanged. and third Figure step 2 shows remained unchanged. Figure 2 shows the folding step of a prototype in configuration 2 (a, b, c) and 4 the folding(d, step e, f).of a prototype in configuration 2 (a, b, c) and 4 (d, e, f).

Figure 2.FigureStep 2. of Step folding of folding in Conﬁgurationin Configuration 22 (a (–ac–) cand) and Configuration Conﬁguration 4 (d–f). 4 (d–f).

Each configurationEach configuration has been has characterizedbeen characterized by meansby means of of the the final final geometrygeometry attained attained after after the the shape recovery. Dueshape torecovery. the softness Due to the of softness the solar of the sail, solar no sail, contact no contact measurement measurement can can be be employed. employed. For For this this purpose, a Next Engine Desktop 3D Scanner has been used. This device was based on purpose, amulti-stripe Next Engine laser triangulation: Desktop 3D a Scannermultiple laser-stripes has been are used. projected This by device four solid was state based lasers onwith multi-stripe a laser triangulation: a multiple laser-stripes are projected by four solid state lasers with a wavelength of 650 nm and is registered through a 3.0 Megapixel CCD image camera. In order to preserve the final geometry, the sail has not been moved according to this procedure. The folded sail was positioned onto a measurement table. A 250 W infrared lamp was placed over the sail and the deployment process was executed (Figure3a, the last folding was the first that unfolds). The lamp was removed and the laser scanner unit was mounted on a supporting frame, which allowed an optical path perpendicular to the sail (Figure3b). This configuration allowed us to acquire the necessary point cloud by a single scan. Actuators 2019, 8, x FOR PEER REVIEW 5 of 10

wavelength of 650 nm and is registered through a 3.0 Megapixel CCD image camera. In order to preserve the final geometry, the sail has not been moved according to this procedure. The folded sail was positioned onto a measurement table. A 250 W infrared lamp was placed over the sail and the deployment process was executed (Figure 3a, the last folding was the first that unfolds). The lamp Actuatorswas2019 removed, 8, 38 and the laser scanner unit was mounted on a supporting frame, which allowed an 5 of 10 optical path perpendicular to the sail (Figure 3b). This configuration allowed us to acquire the necessary point cloud by a single scan. Therefore, no merging operation was needed. It is Therefore, no merging operation was needed. It is well-known that the alignment and merging of data well-known that the alignment and merging of data coming from different scans can affect the final comingaccuracy from since different the change scans of can the a ffenvironmentalect the final accuracylight and the since light the path. change The experimental of the environmental conditions light and thehave light been path. different The from experimental the real ones conditions in the space. have Although been di thefferent aluminum from thefoil realhas a ones thickness in the of space. Although0.012 mm, the aluminum the effect of foil gravity has acould thickness affect ofthe 0.012 opening mm, results. the eff Forect ofthis gravity reason, could the heating affect and the openingthe results.opening For this on Earth reason, have the been heating provided and thewith opening the sail onlaying Earth on havea plane been in order provided to give with symmetry the sail to laying on athe plane system. in order to give symmetry to the system.

FigureFigure 3. 3.Heating Heating ((aa)) andand reverse engineering engineering (b) ( blayouts.) layouts.

TheThe data data acquired acquired through through the the ScanStudio ScanStudio have have been been saved saved and and opened opened in Geomagic in Geomagic Studio. Studio. ThisThis environment environment is dedicatedis dedicated to to the the fullfull processingprocessing of of point point clouds clouds and and general general pre-processing pre-processing operationsoperations have have been been provided: provided: thethe datadata have be beenen cleaned cleaned by by means means of isolated of isolated points points of the of the backgroundbackground and and the originalthe original ﬁle formatfile format has has been been changed changed into into Standard Standard Triangulation Triangulation Language Language (STL). (STL). The full customized elaboration has been designed in Wolfram Mathematica 11. This The full customized elaboration has been designed in Wolfram Mathematica 11. This environment has environment has been chosen for its ability to allow a full object-oriented management of been chosen for its ability to allow a full object-oriented management of mathematical data structures. mathematical data structures. The system used for the elaboration is a laptop with an Intel Core The systemi7-8565U used [email protected], for the elaboration 16GB LPDDR3 is a installed laptop with memory, an Intel nVidia Core GeForce i7-8565U MX150 [email protected], with 2 GB DDR516GB LPDDR3memory. installed Required memory, time: nVidiaopening GeForce of the STL MX150 geometry with 28 2 GB s, graphing DDR5 memory. of the rough Required geometry time: 52 opening s, of therotation STL geometry and translation 28 s, graphingof data with of respect the rough to the geometry regression 52 plane s, rotation 8 s, plot andof height translation deviations of data18 s, with respectcontour to the plot regression of the height plane deviations 8 s, plot 19 of s, height histogram deviations of the height 18 s, contourdeviations plot 0.3 ofs, and the vector height plot deviations of 19 s,the histogram normal 288 of thes. height deviations 0.3 s, and vector plot of the normal 288 s.

3. Results3. Results and and Discussion Discussion TheThe data data structure structure has has been been loaded loaded directly directly into into Mathematica. Mathematica. TheThe STLSTL filefile has been divided into vertexinto and vertex polygon and polygon data for data point for height point height and normal and normal investigations, investigations, respectively. respectively. The The original original points points for the first configuration are reported in Figure 4a. The number of the vertices and triangles for the first configuration are reported in Figure4a. The number of the vertices and triangles have have been 367,390 and 731,371, respectively, which allow a spatial accuracy of about 0.1 mm. Since beenthey 367,390 are referred and 731,371, to the respectively,measurement whichsystem allowpositioning, a spatial a rotation accuracy and of a abouttranslation 0.1 mm. have Sincebeen they are referredapplied according to the measurement to the points system regression positioning, plane. In this a rotation case a slight and a rotation translation along have x (2.27°) been and applied according to the points regression plane. In this case a slight rotation along x (2.27 ) and along y (1.49 ) ◦ ◦ showed a very symmetric deployment of the sails (Figure4b). Similar trends have been obtained for the other configurations. After the alignment, the configurations were the ones reported in Figure5. It is evident that the behaviors are deeply different: the reached maximum height ranged between 11 mm (obtained at the fourth configuration) and 64 mm (at the second one). The representation scales are different to highlight local conditions. The main difficulty lies in the corner opening, which is completely achieved only by rightly combining the wire configuration and folding strategy. Actuators 2019, 8, x FOR PEER REVIEW 6 of 10 along y (1.49°) showed a very symmetric deployment of the sails (Figure 4b). Similar trends have been obtained for the other configurations.

Actuators 2019, 8, x FOR PEER REVIEW 6 of 10

Actuators 2019, 8, 38 6 of 10 along y (1.49°) showed a very symmetric deployment of the sails (Figure 4b). Similar trends have

(a) (b)

Figure 4. Original point for the first configuration (a) and after rotation and translation (b).

After the alignment, the configurations were the ones reported in Figure 5. It is evident that the behaviors are deeply different: the reached maximum height ranged between 11 mm (obtained at the fourth configuration) and 64 mm (at the second one). The representation scales are different to highlight local conditions.(a) The main difficulty lies in the corner opening,(b) which is completely achieved Figureonly by 4. rightlyOriginal combining point for the the ﬁrstfirst wire conﬁgurationconfiguration configuration (a)) andand and afterafter folding rotationrotation strategy. andand translationtranslation ((bb).).

After the alignment, the(a) configurations were the ones reported in( bFigure) 5. It is evident that the behaviors are deeply different: the reached maximum height ranged between 11 mm (obtained at the fourth configuration) and 64 mm (at the second one). The representation scales are different to highlight local conditions. The main difficulty lies in the corner opening, which is completely achieved only by rightly combining the wire configuration and folding strategy.

(a) (b)

FigureFigure 5.5. Open layout of the sails after alignment, respectively, inin ConﬁgurationConfiguration 11 ((aa),), 22 ((bb),), 33 ((cc),), andand 44 ((dd).).

InIn FigureFigure6 ,6, the the contour contour graphs graphs have have been been developed developed by by using using the the same same color color scale. scale. InIn this this representation,representation, itit isis evidentevident thatthat thethe firstfirst threethree configurationsconfigurations lacklack inin cornercorner completecomplete opening:opening: aa longlong tailtail waswas showed showed in thein histograms,the histograms,(c) which which highlighted highlighted the needs the for needs a further(d ) for deployment.a further deployment. Conversely, the lastFigure configuration 5. Open layout was of characterized the sails after alignment, by a distribution respectively, with in aConfiguration main peak at 1 8(a mm), 2 (b height:), 3 (c), and the sail was mainly4 (d). flat with a central subsidence. The four behaviors can be summarized by box and whisker diagrams (Figure7). The second configurationIn Figure has 6, the shown contour the worst graphs condition have been with developed more than by 25% using of the the heights same greatercolor scale. than 28In mm.this Therepresentation, configurations it is 1,evident 3, and that 4 characterized the first three by conf aboutigurations the same lack median in corner (7.2 complete mm, 7.4 mm,opening: and 6a mm,long respectively),tail was showed show deeplyin the dihistograms,fferent configurations: which highlighted only the fourth the needs configuration for a hasfurther been deployment. symmetrical and, for this best outcome, it has been assessed that all the point heights lie in 5 mm. ± Actuators 2019, 8, x FOR PEER REVIEW 7 of 10

Conversely, the last configuration was characterized by a distribution with a main peak at 8 mm height: the sail was mainly flat with a central subsidence. Actuators 2019, 8, x FOR PEER REVIEW 7 of 10

ActuatorsConversely,2019, 8the, 38 last configuration was characterized by a distribution with a main peak at 87 ofmm 10 height: the sail was mainly flat with a central subsidence.

Figure 6. Representation in the same color scale after sails opening: Configurations 1, 2, and 3 lack in complete opening occurs while, in Configuration 4, the sail is mainly flat.

The four behaviors can be summarized by box and whisker diagrams (Figure 7). The second configuration has shown the worst condition with more than 25% of the heights greater than 28 mm. The configurations 1, 3, and 4 characterized by about the same median (7.2 mm, 7.4 mm, and 6 mm, Figurerespectively), 6. Representation show deeply in the different same color configurat scale afterions: sailssail sonly opening: the ConfigurationsConfigurationsfourth configuration 1, 2, and has 3 lack been in symmetrical and, for this best outcome, it has been assessed that all the point heights lie in ±5 mm. complete opening occursoccurs while,while, inin ConfigurationConfiguration 4,4, thethe sailsail isis mainlymainly flat.flat.

FigureFigure 7. The 7. secondThe second configuration configuration has beenhas been the worstthe worst condition condition while while configuration configuration 1, 3,1, and3, and 4, despite 4, beingdespite characterized being characterized by the same by median, the same are median, deeply are di deeplyfferent. different.

Now,Now, the polygon the polygon data havedata have been been used used to understand to understand how thehow normal-to-the-surface the normal-to-the-surface are aareﬀected affected by the deployment. For the i-th triangle, the normal has been calculated according to by the deployment. For the i-th triangle, the normal has been calculated according to Equation (1). Equation (1).

pp⃗ ×pp⃗ n =−−−−→p1ip2i −−−−→p1ip2i (1) nˆ i = |pp⃗××pp⃗| (1) −−−−→p p −−−−→p p Figurewhere 7.p ij Theare secondthe triangle configuration vertices. hasThe been angle| 1ithe α2i worst ×between1i condition2i | the normal while nconfiguration and the axis 1, 3,z canand be4, despiteobtained being by Equation characterized (2). by the same median, are deeply different. where pij are the triangle vertices. The angle αi between the normal nˆ i and the axis z can be obtained by Equation (2). Now, the polygon data have been used to understand how the normal-to-the-surface are nˆ i zˆ affected by the deployment. For the i-thαi triangle= arccos, the· normal has been calculated according (2)to nˆ i zˆ Equation (1). | |·| | where zˆ was the versor of the z axis. In Figure8, the vector ﬁeld plots for the di ﬀerent sails have been reported. In order to take into account the areap ofp the⃗ ×p i-thp triangle,⃗ the distribution of the angles has n = (1) been weighted proportionally to this value, which|pp has⃗ ×p beenp⃗ calculated| by Equation (3). where pij are the triangle vertices. The angle α between the normal n and the axis z can be obtained by Equation (2). −−−−→p p −−−−→p p A = | 1i 2i × 1i 2i| (3) i 2

Actuators 2019, 8, x FOR PEER REVIEW 8 of 10

n ∙z α =arccos (2) |n| ∙ |z| where z was the versor of the z axis. In Figure 8, the vector field plots for the different sails have been reported. In order to take into account the area of the i-th triangle, the distribution of the angles has been weighted proportionally to this value, which has been calculated by Equation (3).

Actuators 2019, 8, 38 |pp⃗ ×pp⃗| 8 of 10 A = (3) 2 The outcomes, reported in the form of histograms in Figure 8, have shown that the first and the The outcomes, reported in the form of histograms in Figure8, have shown that the first and the third configurations have a similar normal distribution with the main peak at about 10° and the 80% third configurations have a similar normal distribution with the main peak at about 10◦ and the 80% peak at more than 28°. A mixed distribution has been observed for the second configuration, which peak at more than 28◦. A mixed distribution has been observed for the second configuration, which claims that a partial deployment occurred. In this case, one half of the distribution has had an angle claims that a partial deployment occurred. In this case, one half of the distribution has had an angle more than 36°. The best result has been obtained for the fourth configuration: 80% of the normal more than 36◦. The best result has been obtained for the fourth configuration: 80% of the normal have an inclination angle less than 16° with a median at 10.5°. Some deviations from the vertical have an inclination angle less than 16◦ with a median at 10.5◦. Some deviations from the vertical direction have been caused by the wires. In the surrounding area, a local pronounced deformation direction have been caused by the wires. In the surrounding area, a local pronounced deformation has has been observed. been observed.

(a) (b)

FigureFigure 8.8. TheThe conﬁgurationsconfigurations 11 ((aa)) andand 33 ((cc)) havehave shownshown similarsimilar normalnormal distributionsdistributions withwith thethe mainmain

peakpeak atat aboutabout 1010°.◦. In configurationconfiguration 2 ((bb),), aa partialpartial deploymentdeployment hashas occurred,occurred, halfhalf ofof thethe distributiondistribution with an angle of more than 36◦. The best result has been obtained for Configuration 4 (d): the 80% of the normal distributions have shown an inclination less than 16◦ with a median at 10.5◦. 4. Conclusions The final goal of this work has been to suggest an alternative method for the solar sail self-deployment based on shape memory alloys. The adoption of shape memory elements replacing standard electromechanical actuators can lead to a simplification of the solar sail systems and an associated weight reduction. Four prototypes have been manufactured and the deployed configurations Actuators 2019, 8, 38 9 of 10 have been analyzed by a Laser Scanner in order to measure the planarity degree and the critical areas. Four configurations of the solar sails, which are different for folding ways and surface reduction, have been analyzed by means of Box and Whisker diagrams, histogram, and level curves. Configuration 1 has shown the lower surface reduction (52%) with a satisfactory planarity degree. In configuration 2, the surface reduction has increased (up to 75%) by changing the number of folding (five instead of three) but with worst planarity degree due to an un-optimized length of some SMA wires. In Configuration 3, the length and diameter of the wires have been optimized to reach a good planarity degree, which is similar to configuration 1 but with greater surface reduction. Lastly, in Configuration 4, with the same number and length of wire as Configuration 3, two of five folding sections have been outwards. The Box and Whisker diagrams have confirmed that, in Configuration 4, much narrower data when compared with the other ones have been detected and a lower data scattering from the regression plane have been associated. In addition, the histogram and the level curves have shown that the reached maximum height has been about 10 mm while, in Configuration 2, the highest difference (50 mm) has been detected. From the analysis of the level curves reported in Figure6, it has been possible to observe that the most critical areas are the vertices of the triangles in correspondence of which the greater height from the regression plane has been found. From this point of view, the optimization of the folding ways, the number and length of the SMA wire has been successful, which allows us to manage a sail configuration with the following features:

ﬁve deployments; 75% surface reduction; good planarity degree after deployment.

When adopting Laser Scanner reverse engineering, it is possible to optimize the sail deployment or study new conﬁgurations minimizing the number of experiments.

Author Contributions: Conceptualization, all the authors. Methodology, all the authors. Investigations, all the authors. Writing, reviewing end editing, all the authors. Funding: This research received no external funding. Acknowledgments: We acknowledge Piero Plini and Benedetto Iacovone for their assistance in the experimental tests. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Tsiolkowsky, K.E. Extension of Man into Outer Space; United Scientific and Technical Presses: Moscow, Russia, 1921. 2. Tsander, F. From a Scientific Heritage, NASA Technical Translation No. TTF-541; NASA: Washington, DC, USA, 1967. 3. About Small Solar Power Sail Demonstrator “IKAROS”. Available online: http://global.jaxa.jp/projects/sat/ ikaros (accessed on 5 March 2019). 4. Ikaros. Available online: https://earth.esa.int/web/eoportal/satellite-missions/i/ikaros (accessed on 5 March 2019). 5. McCurry, J. Space Yacht Ikaros Ready to Cast oﬀ for Far Side of the Sun. Available online: https: //www.theguardian.com/world/2010/may/17/space-yacht-ikaros-japan-venus (accessed on 5 March 2019). 6. Mori, O.; Sawada, H.; Funase, R.; Endo, T.; Morimoto, M.; Yamamoto, T.; Tsuda, T.; Kawakatsu, Y.; Kawaguchi, J. Development of ﬁrst Solar Power Sail Demonstrator–Ikaros. In Proceedings of the 21st International Symposium on Space Flight Dynamics, Toulouse, France, 13 September 2009. 7. Block, J.; Straubel, M.; Wiedemann, M. Ultralight deployable booms for solar sails and other large gossamer structure in space. Acta Astronaut. 2011, 68, 3984–3992. [CrossRef] 8. Johnson, L.; Young, R.; Montgomery, E.; Alhorn, D. Status of solar sail technology within NASA. Adv. Space Res. 2011, 48, 1687–1694. [CrossRef] 9. Fernandez, J.M.; Lappas, V.J.; Daton-Lovett, A.J. Completely stripped solar sail concept using bi-stable reeled composite booms. Acta Astronaut. 2011, 69, 78–85. [CrossRef] 10. Roman Kezera, Y.A. Thickness requirement for solar sail foils. Acta Astronaut. 2009, 65, 507–518. Actuators 2019, 8, 38 10 of 10

11. Dalla Vedova, F.; Henrion, H.; Leipold, M.; Girot, T.; Vaudemont, R.; Belmonte, T.; Fleury, K.; Le Couls, O. The Solar Sail Materials (SSM) Project—Status of activities. Adv. Space Res. 2011, 48, 1922–1926. [CrossRef] 12. Costanza, G.; Tata, M.E.; Libertini, R. Eﬀect of temperature on the mechanical behavior of Ni-Ti Shape Memory Sheets. In Proceedings of the TMS Annual Meeting, Nashville, TN, USA, 14–18 February 2016. 13. Huang, W. On the selection of shape memory alloys for actuators. Mater. Des. 2002, 23, 11–19. [CrossRef] 14. Costanza, G.; Paoloni, S.; Tata, M.E. IR thermography and resistivity investigations on Ni-Ti shape memory alloys. Key Eng. Mater. 2014, 605, 23–26. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).