SMA-Based System for Environmental Sensors Released from ... - MDPI

aerospace Article

SMA-Based System for Environmental Sensors Released from an Unmanned Aerial Vehicle Lorenzo Pellone 1,† , Salvatore Ameduri 2, *,† , Nunzia Favaloro 3,† and Antonio Concilio 2,† 1 2 3

* †

Aeronautics Systems Engineering Department, Centro Italiano Ricerche Aerospaziali, Via Maiorise, Capua 81043, Italy; [email protected] Adaptive Structures Department, Centro Italiano Ricerche Aerospaziali, Via Maiorise, Capua 81043, Italy; [email protected] Space Propulsion Facilities, Centro Italiano Ricerche Aerospaziali, Via Maiorise, Capua 81043, Italy; [email protected] Correspondence: [email protected]; Tel.: +39-823-623-556 These authors contributed equally to this work.

Academic Editor: Rafic Ajaj Received: 10 October 2016; Accepted: 18 January 2017; Published: 24 January 2017

Abstract: In the work at hand, a shape memory alloy (SMA)-based system is presented. The system, conceived for releasing environmental sensors from ground or small unmanned aerial vehicles, UAV (often named UAS, unmanned aerial system), is made of a door, integrated into the bottom of the fuselage, a device distributor, operated by a couple of antagonistic SMA springs, and a kinematic chain, to synchronize the deployment operation with the system movement. On the basis of the specifications (weight, available space, energy supply, sensors size, etc.), the system design was addressed. After having identified the main system characteristics, a representative mock-up was manufactured, featuring the bottom part of the reference fuselage. Functionality tests were performed to prove the system capability to release the sensors; a detailed characterization was finally carried out, mainly finalized at correlating the kinematic chain displacement with the SMA spring temperature and the supplied electrical power. A comparison between theoretical predictions and experimental outcomes showed good agreement. Keywords: shape memory alloy springs; UAV; sensor release

1. Introduction In the recent years, the research and development of unmanned vehicles have gained much attention in the academic and military communities worldwide. Systems like unmanned aircraft, underwater exploiters, satellites, and intelligent robotics are widely investigated as they have potential (dual) applications both in the military and civil domains. They are developed to be capable of working autonomously, without a human pilot. One of the main challenges is the need to deal with extremely varying situations that arise in complicated and uncertain environments, like unexpected obstacles and enemy attacks [1]. UAV (unmanned aerial vehicles) is an acronym for a wide class of aircraft, ranging from a few centimeters to many meters in wingspan, from simple, hand-operated units to high altitude, long endurance systems similar in operation to manned aircraft. In the literature they are classified as RPA (remotely piloted aircraft), AAV (autonomous aircraft vehicles), and so on [2–6]. A UAV is defined as an aerial craft, flying without human crew on-board; it can be remotely controlled or fly autonomously [6,7]. Over the past three decades, the popularity of UAV or UAS (unmanned aerial systems—in 2005, the U.S. Department of Defense, DoD, started defining them as “flying systems”) has kept growing at an unprecedented rate. Among different types, small-scale UAVs are gaining interest and popularity because: Aerospace 2017, 4, 4; doi:10.3390/aerospace4010004

www.mdpi.com/journal/aerospace

Aerospace 2017, 4, 4

2 of 22

Aerospace 4, 4 at an unprecedented rate. Among different types, small-scale UAVs are gaining 2 of 22 has kept 2017, growing interest and popularity because:

Theyare are a powerful tools scientific research to attractive features such lowhigh cost, • • They a powerful tools for for scientific research due due to attractive features such as lowas cost, high maneuverability, and easy maintenance. Significant progress been made various maneuverability, and easy maintenance. Significant progress has has been made ininvarious investigationareas areas(e.g., (e.g.,dynamics dynamicsmodeling, modeling,flight flightcontrol, control,guidance, guidance,and andnavigation); navigation); investigation • • They Theycan canbebeimplemented implementedininmany manydangerous dangerouscivil civilapplications, applications,like likeemergency emergencymonitoring, monitoring, victim andrescue, rescue, aerial filming, geological forecast, assessment, pollution victim search and aerial filming, geological survey,survey, weatherweather forecast, pollution assessment, fireand detection, andmonitoring. radiation monitoring. fire detection, radiation

The of of UAVs hashas been strongly motivated by military applications. Indeed,Indeed, their Thedevelopment development UAVs been strongly motivated by military applications. development dates back to the development of the first aerial torpedoes, almost 95 years ago. During their development dates back to the development of the first aerial torpedoes, almost 95 years ago. World War I (WWI), the Navy andNavy the Army experimented aerial torpedoes and flyingand bombs. During World War Iboth (WWI), both the and the Army experimented aerial torpedoes flying These first examples highlighted two operational problems: for instance, crews had difficulty in bombs. These first examples highlighted two operational problems: for instance, crews had difficulty launching and recovering the UAV; there were several problems in stabilizing them during flight, in launching and recovering the UAV; there were several problems in stabilizing them during flight, and through the theKorean KoreanWar, War, when military services experimented andso soon. on. Efforts Efforts continued continued through when military services experimented withwith UAV UAV with different missions, sensors, and munitions, in attempts to provide strike and with different missions, sensors, and munitions, in attempts to provide strike and reconnaissance reconnaissance servicescommanders. to battlefield commanders. services to battlefield AtAtthose have hindered the evolution ofof thethe UAV. thosetimes, times,numerous numerousobstacles obstacles have hindered the evolution UAV.Firstly, Firstly,technology technology simply was not mature enough for them to become operational. Secondly, lack of service simply was not mature enough for them to become operational. Secondly, lack of servicecooperation cooperation usually usuallyled ledtotofailure. failure.Currently, Currently,the themain mainshowstoppers showstoppersinclude includemostly mostlynon-technical non-technicalissues, issues,such suchasas lack of service enthusiasm, overall cost effectiveness, and competition with other weapons systems lack of service enthusiasm, overall cost effectiveness, and competition with other weapons systems (like and so so on). on). In In this thisframe, frame,small-scale small-scaleUAV UAVseem seeman (likemanned mannedaircraft, aircraft,missiles, missiles, space-based space-based assets, assets, and anideal ideal choice for collecting information zero involved casualties[8],involved [8],implementation at minimal choice for collecting information with zerowith casualties at minimal implementation costs. Moreover, they seem to position into a segment with no real competitors, costs. Moreover, they seem to position into a segment with no real competitors, simply giving a simply giving aThey novelassume service.a They assume unique role to warfare defense, as surveillance for close-range novel service. unique role toa warfare and defense, as and for close-range and surveillance and in reconnaissance in confined battlefields or urban environments. reconnaissance confined battlefields or urban environments. InInthe theframework frameworkofofthe theMilitary MilitaryNational NationalResearch ResearchProgram Program(MNRP), (MNRP),and andthanks thankstotoAerosekur Aerosekur cooperation, CIRA (the Italian Aerospace Research Centre) has carried out a feasibility study cooperation, CIRA (the Italian Aerospace Research Centre) has carried out a feasibility studyofofa a small smallparafoil-equipped parafoil-equippedUAS UASfor forpatrolling, patrolling,intelligence, intelligence,surveillance, surveillance,reconnaissance, reconnaissance,and andtelecom telecom networking. The system is characterized by an excellent transportability and a quick configurability. networking. The system is characterized by an excellent transportability and a quick configurability. ItsItsrelatively relativelysimple simplearchitecture architectureisissketched sketched(Figure (Figure1—left), 1—left),where wherethe thesail sailand andthe thefuselage fuselageoutlines outlines are shown. The target sensor dimensions are also reported (Figure 1—right). The main features are are shown. The target sensor dimensions are also reported (Figure 1—right). The main features are summarized in Table 1: summarized in Table 1:

Figure1.1.UAS UAS ((unmanned aerialsystems) systems)sketch sketch(left); (left);atmosphere atmospheresensor sensordimensions dimensions(right). (right). Figure ((unmanned aerial Table1.1.UAS UASmain mainfeatures. features. Table

Sail Span (m) Sail Span (m) 1.6 1.6

2 Sail Weight (kg) (Wh)(Wh) Overall Weight (kg)Energy Energy Sail Area Area (m (m2)) Overall 2.4 5.0 38.5 2.4 5.0 38.5


3 of 22


3 of 22

The parafoil (or “sail”) is connected to the fuselage by wires. The body hosts the flight control The parafoil (or “sail”) is connected to the fuselage by wires. The body hosts the flight control unit, the power supply, the engine, the payload and the landing gear system. All of the components unit, the power supply, the engine, the payload and the landing gear system. All of the components are designed to keep thethe system mass and to allow allowaacompact compactpackaging packaging and a are designed to keep system mass andsize sizeatataaminimum minimum and and to and rapid deployment on the ground. The target vehicle is conceived to perform missions at low altitude a rapid deployment on the ground. The target vehicle is conceived to perform missions at low altitude (300(300 m atmsea level) endurance(1(1h).h). A typical mission is sketched in 2.Figure 2. It can at sea level)and andlimited limited endurance A typical mission is sketched in Figure It can bring bring different payloads and sensors, depending selected mission fixed mobile targets different payloads and sensors, depending on on thethe selected mission (as (as fixed andand mobile targets identification, border patrol, traffic control,and andmany many others, others, also applications). identification, border patrol, traffic control, alsoinvolving involvingcivil civil applications).

Figure2.2.Typical Typical mission mission of Figure ofaaUAS. UAS.

The device presented in this paper is aimed at storing and releasing sensors from the UAV. In The device presented in this paper is aimed at storing and releasing sensors from the UAV. In this this sense, it belongs to the robotic end effector family [9], being in practice an interface between the sense, it belongs to the robotic end effector family [9], in practice an interface between the UAV UAV and the external environment. With reference to being the standard classification of gripper-releasing and devices the external environment. With referencenegative to the standard of gripper-releasing (mechanical, thermo-mechanical, pressure,classification magnetic, and so on) [10], it candevices be (mechanical, thermo-mechanical, negative pressure, magnetic, and so on) [10], can be identified identified as a thermo-mechanical system in the sense that it is actuated through it shape memory alloy as a thermo-mechanical system inrequires the sense that it is actuated through shape alloy whose (SMA), whose functioning heat supply. Moreover, the device wasmemory conceived to be(SMA), employed for unknown environments. In fact, it is switched on during certain phases of employed flight, but itfor does not functioning requires heat supply. Moreover, the device was conceived to be unknown require any feedback status ofon itsduring task. certain phases of flight, but it does not require any environments. In fact, itonisthe switched SMAs are characterized by the co-existence of two different and distinct states or phases inside feedback on the status of its task. the same physical domain, namely martensite and austenite [11].and Thedistinct switch between states SMAs are characterized by the co-existence of two different states orthese phases inside may be continuous and depends on the stress and temperature levels. Microscopically, these the same physical domain, namely martensite and austenite [11]. The switch between these states may variations are related to a re-arrangement of the crystal structure within the metal while be continuous and depends on the stress and temperature levels. Microscopically, these variations are macroscopically they give rise to relevant deformations, ranging up to several unit percentage almost related to a re-arrangement of the crystal structure within the metal while macroscopically they give rise reaching 10%. Together with this effect, some other material features vary, like the damping to relevant deformations, ranging up1% to several percentage almost reaching 10%.(inTogether with this coefficient (moving from barely to moreunit than 10%) or the Young modulus certain cases effect, some other material features vary, like the damping coefficient (moving from barely 1% to more increasing or decreasing by a factor 3—martensite to austenite, or vice versa). The reported numbers thanrelate 10%)to orNiTi-based the YoungSMA. modulus (in certain cases increasing or decreasing by a factor 3—martensite For other compounds, those qualities may change; nevertheless, they give to austenite, versa). reportedinnumbers relate to NiTi-based For other compounds, those an ideaorofvice their strongThe potentiality generating adaptive structuralSMA. elements. In this article, shape memory springs are used tothey construct a motor fortheir deploying taking ofadaptive both qualities mayalloys change; nevertheless, give an idea of strongsensors, potentiality inadvantage generating the strain recoveryIn(contraction) the memory variable rigidity (stiffening) The non-linear— structural elements. this article,and shape alloys springs arepeculiarity. used to construct a motor for hysteretic nature of the SMA materials and the combined strong dependence on the applied load deploying sensors, taking advantage of both the strain recovery (contraction) and the variable and rigidity temperature, make theThe modeling task very challenging and justify several worksand focusing on the (stiffening) peculiarity. non-linear—hysteretic nature of the the SMA materials the combined formalization of constitutive laws load [12–14] and on the empirical-experimental characterization [15,16]. strong dependence on the applied and temperature, make the modeling task very challenging The described system has been specifically designed to release environmental (temperature, and justify the several works focusing on the formalization of constitutive laws [12–14] and on the pressure, etc.) or pollution sensors, in order to get more environmental information. In commerce, empirical-experimental characterization [15,16]. there is a quantity of such miniaturized devices, developed and applied by industry, academia and The described system been specifically designed to detect release environmental (temperature, individual innovators. Forhas instance, there are devices able to non-methane volatile organic pressure, etc.) or pollution sensors, in order to get more environmental information. In commerce, compounds (NMVOC) emissions, smoke, and toxic gases, [17]. During the mission, the UAV could

there is a quantity of such miniaturized devices, developed and applied by industry, academia and individual innovators. For instance, there are devices able to detect non-methane volatile organic


4 of 22

Aerospace 2017, 4, 4 of 22 compounds (NMVOC) emissions, smoke, and toxic gases, [17]. During the mission, the UAV 4could work like an antenna, receiving and transmitting real-time data to a ground receiving station, or a data work like an antenna, receiving and transmitting real-time data to a ground receiving station, or a store, using a micro data recorder. data store, using a micro data recorder. In order to give the UAS some further capability, a simple, affordable, low-cost deployment In order to give the UAS some further capability, a simple, affordable, low-cost deployment system was specified. This paper deals with the design of an environmental sensor spreader (ESS). system was specified. This paper deals with the design of an environmental sensor spreader (ESS). The concept is shown in Figure 3. It is made of a bottom door, an SMA-based sensor mover and a The concept is shown in Figure 3. It is made of a bottom door, an SMA-based sensor mover and a kinematic chain, devoted to ensure the due synchronization between the movement of the different kinematic chain, devoted to ensure the due synchronization between the movement of the different elements andand in particular thethe panel opening to the theground) ground)and and elements in particular panel opening(from (fromwhich whichthe thesensor sensor is is launched launched to the settler displacement (that pushes the sensor to the door). the settler displacement (that pushes the sensor to the door).

1 2 3 4 5 6 7 8

Base Sensor Guide—Positioner SMA Active Spring (red) SMA Antagonistic Spring (blue) Opening Panel Lever Sensors Sensor box

Figure 3. CAD of the environmental sensor spreader (ESS) and its main elements. Figure 3. CAD of the environmental sensor spreader (ESS) and its main elements.

SMA actuators permit a strong reduction of the system parts and, then, its complexity. In similar SMA actuators permit a strong reduction the system and, then, its complexity. similar applications, a servo actuator opens the panelof[18,19], whileparts another servo actuator grips andIn releases applications, servo actuator opens theSMA panel [18,19], while another actuator grips The andpenalty releases the objectsa[20,21]. The use of a bulk component simplifies theservo overall architecture. the objects [20,21]. The a bulkefficiency SMA component the overall architecture. The penalty that shall be paid foruse theof related reductionsimplifies (1 to 5) is well compensated by the decrease in thatthe shallnumber be paid of for pieces the related efficiency (1 to 5) and is well compensatedweight by the decrease the (a good indexreduction for reliability maintenance), savings, in and compactness. advantage also exploited for other on-board components gears, number of piecesThis (a good indexcan forbe reliability and maintenance), weight savings,like andlanding compactness. andexploited others [22–31]. Thisdeployable advantageantennas can be also for other on-board components like landing gears, deployable aspects, like the power supply, the weight, the size, and the maintenance, play critical antennasDifferent and others [22–31]. roles and must be carefully account the early stage of the the design of this specific end Different aspects, like the taken powerinto supply, thefrom weight, the size, and maintenance, play critical effector. The application of SMA actuators offers some advantages, proved by several studies and roles and must be carefully taken into account from the early stage of the design of this specific end patents [32–35], ranging industrial [36], some to the medical-surgery [37], the aerospace [38], and effector. The application offrom SMAthe actuators offers advantages, proved by several studies and the automotive [39], fields. The compactness and the high energy density favor lighter and solid patents [32–35], ranging from the industrial [36], to the medical-surgery [37], the aerospace [38], and the design solutions for instance compliant with the typical aerospace weight and maintenance automotive [39], fields. The compactness and the high energy density favor lighter and solid design requirements [40]. The large transmittable forces and displacements well meet often conflicting solutions for instance compliant with the typical aerospace weight and maintenance requirements [40]. desiderata, like grasp and deployment specifications, typical of hand-like grippers [32]. The The large transmittable forces and displacements well meet often conflicting desiderata, like grasp and possibility of tuning the electrical signal, used for the SMA heating, makes available a signal for a deployment specifications, typical of hand-like grippers [32]. The possibility of tuning the electrical feedback control, applicable in the low frequency bandwidth [41]. signal, used for other the SMA heating, makes drawbacks available aand signal for a feedback applicable in theThe low On the hand, some general limitations must becontrol, necessarily considered. frequency bandwidth [41]. low mechanical efficiency may limit the use of those actuator systems to small applications [40]; the On theinertia other of hand, some general drawbacks and limitations be necessarily considered. thermal the material may exclude any dynamic applicationmust (practically, ≥2 Hz); the electrical The supply low mechanical efficiency may limit the use of those actuator systems to small applications [40]; imposes the use of adequate thermal and electrical insulation strategies. the thermal inertia of the material may exclude any dynamic application (practically, ≥ 2 Hz); All of these aspects were taken into account during the early stage of the device design. In the electrical supply imposes the use of adequate thermal and electrical strategies. compliance with the drafted characteristics, the SMA actuators size (hereinsulation used in the form of springs) All of these aspects were taken account duringthe thepanel earlyand stage of the design. were contained. The kinematic chain into in charge of opening moving anddevice dropping the sensor waswith optimized to assure an adequate synchronization and to global In compliance the drafted characteristics, themovement SMA actuators size (here used in guarantee the form ofasprings) of the overall system springs supposed to be directly the in weresimplification contained. The kinematic chain architecture. in charge of Finally, openingthe the panel were and moving and dropping contact the calm air contained withinmovement the fuselage volume and, thus, convection sensor was with optimized to assure an adequate synchronization and in to natural guarantee a global regime. Force convection strategies would speed up thethe cooling process, thetorelease task simplification of the overall system architecture. Finally, springs were making supposed be directly faster, but would also penalize the system in terms of power consumption (keeping the springs at a


5 of 22


5 of 22

in contact with the calm air contained within the fuselage volume and, thus, in natural convection regime. convection strategiesand would speed investigations up the cooling aimed process, the release task process faster, lower Force temperature). Numerical empirical atmaking enhancing the cooling but wouldinlets/outlets also penalizewill the be system in terms power consumption (keeping the springs a lowerfor through considered in of future works with the scope of optimizing theatsystem temperature). Numerical and empirical investigations aimed at enhancing the cooling process through specific missions and enhancing the technology readiness level (TRL). inlets/outlets will be considered instarted future works withinvestigation the scope of optimizing the system for based specificon The device development from an about its functionality, missions and enhancing the technology readiness level (TRL). geometrical considerations and multi-body simulations. The model reproduced both the positioner The device development started from an investigation about its functionality, on geometrical translation and the door rotation. A concentrated force simulated the action ofbased an active SMA spring considerations andan multi-body simulations. Thewhose model reaction reproduced the positioner and working against antagonistic SMA spring, wasboth estimated throughtranslation the classic forcethe door rotation. A concentrated force simulated the action of an active SMA spring working against displacement curve, experimentally determined. This empirical approach was considered applicable antoantagonistic SMA spring, whose was sophisticated estimated through the classic force-displacement the device conceptual design task,reaction while more models (Finite Element, FE, integrated curve, determined. This empirical approach considered applicable to the device with experimentally dedicated constitutive laws [42,43]) will be taken into was account for future works, focusing on the conceptual design task, while more sophisticated models (Finite Element, FE, integrated with dedicated optimization of the different functional parameters. By correlating the active spring displacement to constitutive laws [42,43]) will be taken future works, to focusing the optimization of the door rotation, the kinematic frameinto wasaccount set. Thefor force, necessary achieveon a certain configuration, the different functional parameters. By correlating the activethe spring displacement to the door rotation, was calculated by imposing the energy balance between elastic energy stored in the antagonistic the kinematic frame was set. The force, necessary to achieve a certain configuration, was calculated by spring and the work done by the external actions (i.e., the active spring force and the sensors weight imposing energy balance energy storedNumerical in the antagonistic spring and the workto moment,theevaluated with between respect the to elastic the door pivot). simulations were related done by the external actions (i.e., the active spring force and the sensors weight moment, evaluated experimental results, obtained through a test campaign on a functional mock-up (Figure 3). The with respect to the door pivot). Numerical simulations were related to experimental results, obtained experiments allowed validating the theoretical model and verifying the system functionality; through a testthey campaign on a functional (Figure 3). The experiments allowed moreover, permitted drafting amock-up preliminary relationship between the SMAvalidating activation the theoretical model and verifying the system functionality; moreover, they permitted drafting a temperatures and the system kinematics. Finally, the achieved performance was compared to the preliminary relationship between the SMA activation temperatures and the system kinematics. Finally, original requirements. the achieved performance was compared to the original requirements. 2. Requirements and Specifications 2. Requirements and Specifications Referring to the baseline mission scenario “detection and surveillance” (Figure 2), where there Referring to the baselinethe mission “detection and surveillance” (Figurewere 2), where there is a long cruise segment, main scenario specifications of the sensor release system related to is the a long cruise segment, the main specifications of the sensor release system were related to the overall overall size and weight of the hosting UAS. According to the European Cooperation Space size and weight of the hosting According to the Standardization, Standardization, ECSS, SpaceUAS. Standard, in detail in European the SystemCooperation EngineeringSpace General Requirements, ECSS, Space Standard, in detail in the System Engineering General Requirements, ECSS-E-ST-10C, ECSS-E-ST-10C, and the Technical Requirements Specification, ECSS-E-ST-10-06C, all ofand the the Technical Requirements Specification, ECSS-E-ST-10-06C, all most of the specifications associated specifications were associated with some attributes, the important of were which was the with some attributes, the words, most important of which was the “traceability”. In generator other words, each single “traceability”. In other each single requisite (son) was linked to its (parent) inside a requisite (son) to the its generator inside a specified set. enables the derivation of specified set.was Thislinked enables derivation(parent) of a “tree”, demonstrating theThis global coherence of the issues a “tree”, demonstrating the global coherence of the issues specs. A synthetic block diagram is reported specs. A synthetic block diagram is reported in Figure 4. in Figure 4.

Figure 4. 4. Requirement tree. Figure Requirement tree.

In the studied case, the high-level requirements are following reported. Above all, the system dimensions should have fit with the fuselage size and allow transporting and deploying the


6 of 22

In the studied case, the high-level requirements are following reported. Above all, the system dimensions should have fit with the fuselage size and allow transporting and deploying the maximum possible number of sensors. The system weight should have been compatible with the maximum payload weight defined for the specific UAS and the related mission. Starting from that, the following constraints were defined (sensor size was fixed to 10 × 2 × 4 mm):

• • • •

System dimensions: ≤100 × 30 × 54 mm System weight (without sensors): ≤70 g Total system weight (including sensor): ≤300 g Max power available: ≤15 W

3. Working Principle of the Environmental Sensor Spreader (ESS) The equilibrium state between the two SMA springs, in turn modulated by their excitation levels, creates a certain ESS system configuration. Its mechanics follows these steps: Step 1—“Initiation”: In this phase, the two SMA springs are in a pure elastic equilibrium (there is no activation by any external source). At their installation, the springs are stretched and connected each other, thus achieving the pre-load equilibrium. Their stretching induces the transformation of the initial full austenite into martensite. To this aim, a material was selected, characterized by a martensite start temperature (Ms ) very close to the specific environmental conditions. Therefore, the necessary stress increment to induce the desired transformation may be kept at a minimum. On the other hand, it should be mentioned that this limit cannot be fully approached in order to avoid the insurgence of non-commended activations due, for instance, to a hot sunny day. The martensite fraction, ξ, produced during the stretching can be expressed as a function of the curvilinear abscissa along the experimental black curve in Figure 5. There, it shows the force-displacement history the springs undergo when they are integrated into the system. The initial linear slope of the curve corresponds to a classical material behavior, regulated by the austenite Young modulus. As the applied load arises, the gradual transformation into martensite occurs. The transformation process may continue up the phase transformation is completed. In the graph, this is evidenced by a relevant change of the line angle, characteristic of this transitional effect. The higher the martensite fraction in the pre-load condition, the wider the recoverable displacement through the activation. In the pre-load condition, the sensor is ready in the dispenser (guide-positioner, Figure 6—top). Considering the relation among the rigidity, K, of the springs and the wire diameter, di, the winding diameter, D, the number of windings, N, and the shear modulus, G: K=

1 di4 G 8 D3 N

(1)

and adapting the shear—normal elastic moduli relation to the ξ weighted average of the martensite, Em , and austenite, Ea , Young moduli: G=

E Em ξ + Ea (1 − ξ) = 2(1 + ν ) 2(1 + ν )

(2)

Aerospace Aerospace2017, 2017,4,4,4 4

77ofof2222

Figure Figure5.1.System Systemworking workingpoints; points;force forcedisplacement displacementcurve curvefor forthe thenon-activated non-activatedSMA SMAspring spring(black); (black); force forcedisplacement displacementcurve curvefor forthe thefully fullyactivated activatedSMA SMAspring spring(red); (red);intermediate intermediateactivation activationcurve curve (dashed); elastic response of the system (blue) (dashed); elastic response of the system (blue)

Anestimate estimateof ofthe themartensite martensitefraction fractioncan canbe beobtained obtainedat ateach eachpoint pointof ofthe thecurve. curve.In Inparticular, particular, An ξ results equal to the 85.5% in pre-load condition. This verification was experimentally performed by ξ results equal to the 85.5% in pre-load condition. This verification was experimentally performed by releasingthe thespring springafter afterhaving havingreached reachedthe thetarget targetelongation. elongation. ItItmoves movesthen thenalong alongaastraight straightline, line, releasing theslope slopeofofwhich whichis is related “K”, Equation using Equations (1) and (2),estimate an estimate the related to to “K”, Equation (1).(1). By By using thethe Equations (1) and (2), an of theof the effectively achieved ξ is easily produced. effectively achieved ξ is easily produced. Step2—“Door 2—“Dooropening”: opening”:ItItisismade madeofoftwo twocoordinated coordinatedmovements: movements:aatranslation translationofofthe thesensors sensors Step guidecoupled coupledtotoaarotation rotationofofthe theejection ejectionpanel. panel.InInparticular: particular: guide a. a.

At first, the active SMA spring (red in Figure 6) is excited and pulls the sensors guide from At first, the active SMA spring (red in Figure 6) is excited and pulls the sensors guide from position 1 to 2 (Figure 6—middle); position 1 to 2 (Figure 6—middle); b. At position 2, while the sensor guide continues its movement, a lever system (Figure 3, At position 2, while the sensor guide continues its movement, a lever system (Figure 3, b. component (6) opens the door and allows releasing the sensor (position 3, Figure 6—bottom). component (6) opens the door and allows releasing the sensor (position 3, Figure 6—bottom). Step 3—“Door closing”: Once the sensor is dropped, the active SMA spring excitation is turned Stepthe 3—“Door closing”: Once the sensor dropped, active in SMA is turned off and antagonist elastic element (blue)isreturns thethe system its spring originalexcitation configuration; the off and themay antagonist elastic element (blue) returns the system in its configuration; process be accelerated by activating this other spring, inoriginal turn. The switch railthe is process driving may be accelerated this other spring, in turn. The switch rail is driving backward to close backward to close by theactivating door. the door.

Aerospace 2017, 4, 4 Aerospace 2017, 4, 4

8 of 22 8 of 22

Figure 6. ESS working principle schematic, highlighting the main phases of the sensor’s release. Figure 6. ESS working principle schematic, highlighting the main phases of the sensor’s release.

4. 4. Modeling Modeling The Therein, the the “active” “active” The system system was was modelled modelled according according to to the the schematic schematic reported reported in in Figure Figure 7. 7. Therein, red and the “antagonistic” blue springs are shown as a joint in B. The positioner is not sketched; red and the “antagonistic” blue springs are shown as a joint in B. The positioner is not sketched; itit is is linked thethe point B. B. AsAs thethe redred spring contracts, it moves to the linked to tothe thetwo twoelastic elasticelements elementsthrough through point spring contracts, it moves to left, the within a slidea (grey the picture). The green AD represents a lever that is rigidly left, within slide in (grey in the picture). Thesegment green segment AD represents a lever thatconnected is rigidly to the door, schematized as a further green line from A to the left. This couple is hinged in is A,hinged so thatin a connected to the door, schematized as a further green line from A to the left. This couple motion of the point D causes its rotation. The first picture, (a), represents the initial state of the system: A, so that a motion of the point D causes its rotation. The first picture, (a), represents the initial state aofclosed door, non-excited and non-excited equilibrated and springs, with point B at thewith rightpoint slideBendpoint. the system: a closed door, equilibrated springs, at the right slide

endpoint.


9 of 22 9 of 22

Figure7.7.Schematic Schematiclayout layoutofofthe theproposed proposeddevice: device:initial initialstate stateofofthe thesystem system(a); (a);leftward leftwardtranslation translation Figure of the point B due to the spring activation (b); rotation of the door (c). of the point B due to the spring activation (b); rotation of the door (c).

The Theactivation activationofofthe thered redspring springproduces producesan aninitial initialtranslation translationofofthe thepoint pointB,B,which whichmoves movestoto the theleft leftslide slideendpoint endpoint(Figure (Figure7b). 7b).Since Sincethis thismoment, moment,aafurther furthercontraction contractionofofthe thered redspring springforces forces the themotion motionofofthe theslide slidetotothe theleft; left;as asititisisrigidly rigidlyconnected connectedtotothe thepoint pointDD(grey (greyline), line),this thiscauses causesaa rotation rotationofofthe thelever lever(segment (segmentAD) AD)and andthe thedoor door(Figure (Figure7c). 7c).ItItisisimportant importanttotoremark remarkthat thatthe theblue blue spring springand andthe thegrey greyconnection connectionare arenot notinterfering interferingeach eachother; other;they theyare arejust justoverlapped overlappedininthe the1D 1D drawing drawingbut butthey theyhave haveno nopoint pointinincommon. common.The Themovement movementmay maycontinue continueuntil untilthe theleft leftspring springreaches reaches its itsmax maxcontraction. contraction.AAblock blockdid, did,however, however,limit limitthe thedoor doorrotation rotationatat75° 75◦from fromthe theinitial initialvalue valueθθ0,0 , following followingthe thespecifications. specifications.In InTables Tables22and and3,3,the themain mainSMA SMAspring springparameters parametersand andthe themain mainoverall overall system systemfeatures featuresare arereported, reported,ininorder. order. The Theselected selectedNiTiNol NiTiNolalloy alloy(Ni (Ni49.8%) 49.8%)was wasexperimentally experimentally characterized. characterized.Austenite Austeniteand andmartensite martensiteYoung Youngmoduli moduliwere weredetermined determinedthrough throughtensile tensiletests testsand and Differential DifferentialScanning ScanningCalorimetry, Calorimetry,DSC, DSC,measurements measurementsatat“0” “0”load loadcondition conditionafter afterthe thenecessary necessary ◦ C), stabilization stabilizationcycles. cycles.The Thematerial materialunderwent underwenttens tensofofmechanical mechanicalcycles cyclesatatroom roomtemperature temperature(23 (23°C), and andthe thelast lastmost mostsignificant significantcycles, cycles,in interms termsof ofconvergence, convergence,are areplotted plottedin inFigure Figure8a. 8a. Table Table2.2.SMA SMAspring springcharacteristics characteristics[31]. [31]. HelixHelix WireWire DiameterDiameter Diameter Diameter (mm) (mm) (mm) (mm) 5.5

5.5

0.5

0.5

Windings Windings

Martensite Transformation Austenite Austenite Martensite Transformation Unloaded Unloaded Young Young Temperatures Young Young Temperatures Density Length, Density (kg/m3 ) Length,Material L. Material Modulus Modulus Modulus Modulus As, Af, Ms,As, Mf,Af, Ms, Mf, (kg/m3) L. (mm) ◦ (GPa) (GPa) Rs, Rf ( C) (mm) (GPa)

23

23

25.0

25.0

NiTiNOL

103.3

NiTiNOL

57.7

103.3

(GPa)

Rs, Rf (°C)

65.9; 78.4; 36.4; 65.9; 78.4; 36.4;6450 −3.8; 73.7; 49.8 57.7

−3.8; 73.7; 49.8

6450


10 of 22


10 of 22

Table 3. Main system features. Table 3. Main system features. Device Device Shape Shape (mm)

(mm)

1515 × ×1515××33

Device

Device CoG

AC Vertical

BC Length, a,

Slot Length, d,

Initial Angle, Initial Angle, θ0, (°)θ0 , (◦ )

Device BC Length, a, Slot Length, WeightWeight DistanceDevice to A CoGDistanceAC Vertical (mm) (mm) d, (mm) (g) Distance to A (mm) Distance (mm) (mm)

(g) 10.010.0

(mm) 7.5

(mm) 5.0

7.5

5.0

70.0

70.0

(a)

4.0

4.0

45.6 45.6

(b)

Figure 8. SMA spring material characterization: tensile cycles (a); and Differential Scanning

Figure 8. SMA spring material characterization: tensile cycles (a); and Differential Scanning Calorimetry, DSC, measurements (b). Calorimetry, DSC, measurements (b).

The modelling started with the formalization of the non-linear relation between the door The modelling started with the formalization of thebetween non-linear relation doorline), rotation, rotation, θ–θ0 (respectively, current and initial angles the lever ACbetween and the the ground the point B translation. These are the main descriptors of the kinematic chain. For displacements θ–θand (respectively, current and initial angles between the lever AC and the ground line), and the point 0 that were lower the main slide length, d, Figure 7b,kinematic the angle chain. was kept at θ0 (closed door).lower B translation. Thesethan are the descriptors of the Forconstant displacements that were displacements than7b, d, athe trialangle valuewas is used toconstant start the iteration to assess the For running angle. thanFor the slide length,larger d, Figure kept at θ0 (closed door). displacements The equation linking ψ, the angle between the slide and the ground line, and θ can be written as: larger than d, a trial value is used to start the iteration to assess the running angle. The equation linking

ψ, the angle between the slide and the ground line,θand θ can be written as: sin

ψ = sin

−

(3) b θ h ψ = sin−1 sin − (3) In the above formula, b, a, and h represent a the length a of the a segments AD (lever), BD (slider

system), and the vertical distance between the point C and A, respectively. The relation between the In the above formula, b, a, h represent the the length of θthe longitudinal displacement ofand the point B, xB, and angle is,segments instead: AD (lever), BD (slider system),

and the vertical distance between the point C and A, respectively. The relation between the longitudinal (4) θ −θ is, instead: θ − = ψ the − angle displacement of the point B, xB , and The sign of the estimated residual, R, derives from the bisection method that generates a further a cos ψ(3) − band θ −until cos θ [cos(4) 0 ] − x B = Roccurs (R  0). The resulting (4) value for θ used again in Equations convergence rotation—displacement curve is reported in Figure 9, for θ ranging between 0° and 90°. Spring The sign ofis the estimated residual, R, derives length from (zero the bisection method that generates displacement plotted with respect to its equilibrium displacement corresponds then a further value used again Equations andin (4) convergence occurs (R → 0). to x = 54 mm).forA θcircular markerinhighlights this(3) state the until picture. Moving from this point, of B do not produce any rotation until they lower9,than After that, the door0◦starts Thedisplacements resulting rotation—displacement curve is reported inare Figure for θd. ranging between and 90◦ . opening and the curve leaves the x axis. Theto spring activation force was(zero computed by expressing the Spring displacement is plotted with respect its equilibrium length displacement corresponds system energy balance, Equation (5). It takes into account the elastic energy stored in the antagonistic then to x = 54 mm). A circular marker highlights this state in the picture. Moving from this point, spring, left side, spring work, first term on they the right side, and thed.work by thestarts displacements of B the do active not produce any rotation until are lower than Afterperformed that, the door door weight momentum around A, second term on the right side:

opening and the curve leaves the x axis. The spring activation force was computed by expressing the system energy balance, Equation (5). It takes into account the elastic energy stored in the antagonistic spring, left side, the active spring first term right ̅ work, ̅= ̅ + on thecos θ −side, θ θand the work performed (5)by the door weight momentum around A, second term on the right side: Zs 0

Fantag (s)ds =

Zs 0

Fact ds + Wrcg

θ Z(s)

cos θ − θ0 dθ

θ0

(5)


11 of 22 11 of 22

theequation equationabove, above,W Wand andrrcgcg are InInthe are the the door door weight weightand andthe thedistance distanceofofits itscenter centerofofgravity gravityfrom from thepivot, pivot,ininorder. order.The Thecosine cosinefunction function expresses expresses the the variation variation of the of the the weight weightmomentum momentumarm armwith with respect to A, as the door rotates. F act represents the activation force, unknown, while Fantag relates the respect to A, as the door rotates. Fact represents the activation force, unknown, while Fantag relates the displacementofofthe theantagonistic antagonisticspring springtoto reaction. This relation between applied force and displacement itsits reaction. This relation between thethe applied force and the the induced displacement is experimentally determined by measuring the spring displacement induced displacement is experimentally determined by measuring the spring displacement caused by caused by different loads. different loads.

Figure9.9.Door Door rotation rotation vs. vs. super elastic Figure elastic (SE) (SE) spring springdisplacement. displacement.

Equation (5) can be numerically solved for Fact, in order to relate this parameter to the induced Equation (5) can be numerically solved for Fact , in order to relate this parameter to the induced displacement, xB. The characteristic system working points (representing the non-activated and the displacement, xB . The characteristic system working points (representing the non-activated and the fully-activated states) can be found by displaying three different curves onto a single graph, Figure fully-activated states) can be found by displaying three different curves onto a single graph, Figure 5: 5:

• • The ” curve curve(blue); (blue); Thejust-mentioned just-mentioned“F “Fact act-xB B” • • The SMAspring; spring;this thiscurve curvewas wasplotted plotted Theforce forcedisplacement displacementcurve curve (black) (black) for the non-activated non-activated SMA theenvironmental environmental temperature, temperature, TTenv : :asasthe rises, thethe transformation intointo martensite is atatthe theload load rises, transformation martensite env isenforced; enforced; Theforce forcedisplacement displacementcurve curve(red) (red) for for the the fully fully activated • • The activated SMA SMAspring; spring;this thiscurve curverefers referstotoa apure pure austenitephase phaseand andititwas wasplotted plottedunder underthe theassumption assumptionthe thetemperature temperaturewas wassufficiently sufficientlyhigh highto austenite to keep austenite condition within entire load range. keep purepure the the austenite condition within thethe entire load range. Moving from the pre-load point on the black curve (“equilibrium in no load condition”), any Moving from the pre-load point on the black curve (“equilibrium in no load condition”), any rise of the SMA temperature enforces its transformation into austenite, passing through intermediate rise of the SMA temperature enforces its transformation into austenite, passing through intermediate configurations, like the one depicted by the grey dashed curve. The intersections between the blue configurations, likerepresent the one depicted byfinal the grey dashedpoints curve.ofThe between thepreblue and these curves initial and equilibrium theintersections global system. Since the and these curvesisrepresent initial final slope equilibrium of the global Since the pre-load load condition located on the and constant part of points the black curve, it issystem. possible to assume that condition is located on the constant slope part of the black curve, it is possible to assume that the spring works, moving from a full martensite condition. Full activation may lead to an the 85° spring door works, moving from athan full the martensite may an 85◦ doormargins. rotation, well rotation, well larger requiredcondition. one (75°).Full The activation device then haslead goodtooperational ◦ larger The thanselected the required one (75 allowed ). The device then hassimplification good operational configuration a remarkable of themargins. overall modelling process. The selected configuration allowed a remarkable simplification ofstate, the overall modelling process. In fact, once the two-spring system has been brought to its equilibrium the working philosophy Inmoves fact, once the two-spring system has beenalong brought itscurve equilibrium state, thein working the operational point up and down the to blue that is reported Figure 5philosophy and was moves the operational point (indeed, up and down alongare thea blue curve that is reported Figure 5 below). and was experimentally determined the things bit more complex as will bein explained experimentally determined (indeed, the things are a bit more complex as will be explained below). The active spring cycles between a partial martensite (point A, Figure 10) and a full austenite state as The active spring cycles between a partial martensite (point A, Figure 10) and a full austenite state as


12 of 22 12 of 22

its temperature grows (point B, Figure 10). At the cooling, the active spring mutates again and comes its temperature grows (point B, Figure 10). At the cooling, the active spring mutates again and comes back to a partial martensite phase. back to a partial martensite phase. This description is strictly valid if a classical antagonistic elastic spring is considered. Since this This description is strictly valid if a classical antagonistic elastic spring is considered. Since this last element is also made of the same shape memory alloy, after having moved from the initial point last element is also made of the same shape memory alloy, after having moved from the initial point A A to the working point B, the forced reverse transformation does not occur according to the blue to the working point B, the forced reverse transformation does not occur according to the blue curve, curve, Figure 5. Instead, because austenite is created, the return happens along a different, more Figure 5. Instead, because austenite is created, the return happens along a different, more inclined inclined slope, Figure 10 (B  C). As this other component cools, however, the working point moves slope, Figure 10 (B → C). As this other component cools, however, the working point moves up again up again to the initial one (C  A). to the initial one (C → A). Furthermore, the segments BC and CA should not be considered as straight, but as part of Furthermore, the segments BC and CA should not be considered as straight, but as part of curves, curves, characteristic of the SMA change of phase. This takes place in theory. In the practical characteristic of the SMA change of phase. This takes place in theory. In the practical application, from application, from one side the two segments are practically coincident and, on the other side, this one side the two segments are practically coincident and, on the other side, this event does not add event does not add any particular information. Then, for the scope of the present work, this behavior any particular information. Then, for the scope of the present work, this behavior complication may be complication may be neglected (i.e., segments AB and BC are hardly distinguished and may be neglected (i.e., segments AB and BC are hardly distinguished and may be considered straight) and the considered straight) and the system may be described by the curves already illustrated in Figure 5. system may be described by the curves already illustrated in Figure 5. For the reader’s benefit, an ordinated description of Figure 10 is reported. The equilibrium state, For the reader’s benefit, an ordinated description of Figure 10 is reported. The equilibrium A, is reached at temperature T1. The active spring is then warmed to temperature T3 (T2 is an state, A, is reached at temperature T1. The active spring is then warmed to temperature T3 (T2 is an intermediate value). The working point then moves from A to B. When the active spring is cooled intermediate value). The working point then moves from A to B. When the active spring is cooled and and the antagonistic spring is excited, the working point moves from B to C, according to the partial the antagonistic spring is excited, the working point moves from B to C, according to the partial to to full austenitic characteristic (curve BC). As the antagonistic spring also cools, the working point full austenitic characteristic (curve BC). As the antagonistic spring also cools, the working point again again approaches the original point A. AB and BC are not namely straight lines but they are approaches the original point A. AB and BC are not namely straight lines but they are represented in represented in that way for the sake of clarity. that way for the sake of clarity.

Figure 10. Theoretical behavior of a two-spring two-spring SMA SMA system. system.

5. Manufacture and Test Test 11). A dedicated experimental campaign was A conceptual conceptual prototype prototypewas wasthen thenrealized realized(Figure (Figure 11). A dedicated experimental campaign organized withwith the double aim aim of appreciating the real system ability in deploying the sensors and was organized the double of appreciating the real system ability in deploying the sensors monitoring the the main device parameters like thethespring and monitoring main device parameters like springtemperature temperatureand anddisplacement, displacement, the door rotation, the the supplied supplied energy, energy,and andso so on. on. In In particular, particular, two two different differenttest testcampaigns campaignswere wereperformed: performed: rotation, • • • •

Functionality tests, tests, aimed aimed at at verifying verifying the the system system operation; operation; Functionality Monitoring tests, for characterizing the device state and correlating the various descriptors. Monitoring tests, for characterizing the device state and correlating the various descriptors.


13 of 22


13 of 22


13 of 22 13 of 22

Figure 11. Prototype main components.

Figure 11. Prototype main components.

Pictures of the setup block diagram and the experimental setup are shown in Figures 12 and 13, Figure 11. Prototype main components. respectively. Thesetup following was used: Pictures of the blockinstrumentation diagram and the experimental setup are shown in Figures 12 and 13, Figure 11. Prototype main components. respectively. The following instrumentation was used: • The prototype of the sensor release system, Figure 11, clamped on a supporting structure;

• • • • • • •

Pictures of the setup block diagram and the experimental setup are shown in Figures 12 and 13, Pictures of the setup block diagram and thefor experimental setup shown in Figures 12 and 13, • prototype A four channel, 24sensor MHz signal generator, the generation of are the signals, necessary The the release system, 11, clamped onlogical a supporting structure;to respectively. Theoffollowing instrumentation wasFigure used: respectively. following instrumentation was springs; used: drive theThe power suppliers that fed the SMA A four channel, 24ofMHz signal generator, for the 11 generation of athe logical signals, necessary to • the sensor release system, , clamped on supporting structure; • The Twoprototype power supplies, operating within 0–30Figure V and11 0–10 A current • The prototype of the sensor release system, Figure , clamped on aranges; supporting structure; drive the power suppliers that fed the SMA springs; • channel, 24 MHz signal generator, for the generation of the logical signals, necessary to • A Anfour oscilloscope, dedicated to monitor the driving logical signals; • A four channel, 24 MHz signal generator, for the generation of the logical signals, necessary to thesupplies, powersystem, suppliers thatwithin fed and the 0–30 SMA springs; Two power operating V and 0–10 A current ranges; • drive An acquisition to manage store the output analogic signals (50 ms sample rate); drive the power suppliers that fed the SMA springs; • Two power supplies, operating within 0–30 V and 0–10 A current ranges; • oscilloscope, Two power “K-type” thermocouples; An dedicated to monitor driving signals;ranges; • Two supplies, operating withinthe 0–30 V and logical 0–10 A current • An oscilloscope, dedicated to monitor the driving logical signals; • acquisition A digital camera to take pictures of thestore system, the experiments. • An oscilloscope, dedicated to monitor the driving logical signals; An system, to manage and the during output analogic signals (50 ms sample rate); • An acquisition system, to manage and store the output analogic signals (50 ms sample rate); • An acquisition system, to manage and store the output analogic signals (50 ms sample rate); TwoTwo “K-type” thermocouples; • “K-type” thermocouples; • Two “K-type” thermocouples; A digital camera to take pictures duringthe theexperiments. experiments. • A digital camera to take picturesofofthe thesystem, system, during • A digital camera to take pictures of the system, during the experiments.

Figure 12. Setup block diagram.

Figure12. 12. Setup block Figure blockdiagram. diagram. Figure 12.Setup Setup block diagram.

Figure 13. Experimental setup.

Figure 13. Experimental setup.

Figure13. 13. Experimental Experimental setup. Figure setup.


14 of 22


14 of 22

Both the functionality and the monitoring tests were performed by activating both of the springs, alternately. The operationsand were follows. Two anti-phase logical step signals were Both the functionality thestructured monitoringas tests were performed by activating both of the springs, alternately. The operations structured assprings, follows. separately. Two anti-phase logical step suppliers signals were produced to drive the active were and antagonistic The two power were produced to drive the(Figure active and springs, separately. The two power suppliers were thenthe then activated in turn 14).antagonistic Each complete cycle (activation—deactivation) took 100 s and activated turn (Figure complete (activation—deactivation) tooka100 s andofthe springs wereinheated for 5 s. 14). TheEach power suppliescycle provided a current of 2.5 A under voltage 1.2 V, springs were heated for 5 s. The power supplies provided a current of 2.5 A under a voltage of 1.2 V, corresponding to 3 W per spring (Direct Current, DC). corresponding to 3 W per spring (Direct Current, DC). An example of the performed functionality tests is shown in Figure 15. Therein, a sequence of An example performed testsrelease. is shown Figure Therein, a sequence of video shots details of thethe door openingfunctionality and the sensor Atinfirst, the15. sensor moves horizontally, video shots details the door opening and the sensor release. At first, the sensor moves horizontally, pushed by the positioner (first frame: start of the cycle; second frame: positioner moved and door pushed by the positioner (first frame: start of the cycle; second frame: positioner moved and door ready for rotation). Then, the door starts opening (third and fourth frames: intermediate door rotation ready for rotation). Then, the door starts opening (third and fourth frames: intermediate door rotation and start of the sensor drop). Finally, the sensor is released (fifth frame). The system operability and start of the sensor drop). Finally, the sensor is released (fifth frame). The system operability was was then proved. The vertical dashed line indicates the initial location of the positioner. A system then proved. The vertical dashed line indicates the initial location of the positioner. A system characterization was then carried out for evaluating the main working parameters. The system was characterization was then carried out for evaluating the main working parameters. The system was driven for several cycles. Two thermocouples monitored the spring temperature, while the camera driven for several cycles. Two thermocouples monitored the spring temperature, while the camera captured different In Figure Figure16, 16,the theactivation activationsignal, signal, the temperature captured differentstates statesofofthe theworking working device. device. In the temperature and thethe corresponding along aa complete completeactivation-deactivation activation-deactivation cycle and correspondingdisplacement displacement are are plotted plotted along cycle forfor both the SMA elements over a period of 1.8 min. both the SMA elements over a period of 1.8 min. The complete two segments, segments,each each0.9 0.9min minlong. long. The first is related The completeoperation operationmay maybe bedivided divided into into two The first is related to to thethe contraction subsequentfreezing freezingininthis thiscondition condition (door contractionofofthe theleft leftspring spring and and its its following following subsequent (door opening—pink the contraction contractionofofthe theantagonistic antagonistic SMA that brings opening—pinkarea); area);the thesecond secondrefers refers instead instead to the SMA that brings thethe device back toto the initial device back the initialconfiguration configuration(door (doorclosing—light closing—lightblue bluearea). area). In In detail, detail, the temperature of ◦ C82(Figure thespring left spring the environmental condition °C (Figure 16—top). A certain theofleft arisesarises from from the environmental condition up toup 82to 16—top). A certain inertia of the powercould supply be observed for around 6 s, since the temperature spring temperature continued of inertia the power supply becould observed for around 6 s, since the spring continued growing growing even it was switched off (dashed lines, Figure 16—top). The contraction started even after it wasafter switched off (dashed vertical vertical lines, Figure 16—top). The contraction started at the at the early supply start, as shown by the corresponding displacement curve (black): from 6.30 6.35 early supply start, as shown by the corresponding displacement curve (black): from 6.30 to to 6.35 min, the length of the left spring passed 10.1 1.0 mm. A specular behavior was observedfor forthe themin, length of the left spring passed fromfrom 10.1 to 1.0tomm. A specular behavior was observed theelastic right elastic element, 16—bottom, they are connected to other each other theshrinks, one right element, FigureFigure 16—bottom, since since they are connected to each and asand theasone shrinks, the other stretches. the other stretches.

Figure14. 14.Logical Logical signals signals to Figure to drive drivethe thepower powersuppliers. suppliers.

The sensor release is accomplished as the active SMA fully shrivels (“spring recovery” label, The sensor release is accomplished as the active SMA fully shrivels (“spring recovery” label, Figure 16—top). Then, it remains for a while in this configuration, as indicated by the “spring stay Figure 16—top). Then, it remains for a while in this configuration, as indicated by the “spring stay contracted” label, up to the end of the pink area. In this phase, the door stays open while the active contracted” label, up to the end of the pink When area. it Inreaches this phase, stays open while the active spring cools in a natural convection regime. 40 °Cthe the door antagonistic device is activated.


15 of 22


15 of 22

spring in a natural it reaches 40 ◦ C the device is activated. This cools temperature was convection considered regime. a good When compromise between twoantagonistic opposite requirements. The This temperature was considered a goodcycle compromise twodemanded opposite requirements. The necessity necessity of starting the deactivation as soon between as possible a rapid activation of the element to reverse the process. Onas the other hand, the lower the austenite percentage is within of other starting the deactivation cycle as soon possible demanded a rapid activation of the other element left spring (and, thus, its temperature), the lower is the force, necessary to reset the system. In spring this to the reverse the process. On the other hand, the the austenite percentage is within the left second the door closes theispositioner back. thisthe point, the system readyphase, for (and, thus,phase, its temperature), the and lower the force, moves necessary to At reset system. In this is second another operation. the door closes and the positioner moves back. At this point, the system is ready for another operation.

Figure 15. Photograph sequence of sensor release. Figure 15. Photograph sequence of sensor release.

Time diagram, measureofofthe theprocess process repeatability. During experiment, Time diagram,Figure Figure17, 17, gives gives a measure repeatability. During the the experiment, a ◦ max negative negative angle angle of 76° predictions and thethe a max 76 was was found, found,in ingood goodagreement agreementwith withthe thenumerical numerical predictions and specifications. The same cycles reported in Figure 18,the ondisplacement-rotation the displacement-rotation A specifications. The same cycles are are reported in Figure 18, on plane.plane. A plateau ranging from about 0 to 4 mm is evident at the top of the graph, corresponding to the phase when

Aerospace 2017, Aerospace 2017, 4, 44, 4 Aerospace 2017, 4, 4

16 of 1622 of 22 16 of 22

plateau ranging from about 0 to 4 mm is evident at the top of the graph, corresponding to the phase thewhen doorthe remains closed. Both the positioner translation the total cycle extension in good door remains closed. Both the translation and the total cycle extension in plateau ranging from about 0 to 4 mm is positioner evident at the topand of the graph, corresponding toare theare phase good agreement with the numerical predictions, previously shown in Figure 9 (abscissa distance agreement with the numerical predictions, previously shown in Figure 9 (abscissa distance between when the door remains closed. Both the positioner translation and the total cycle extension are in the “initial equilibrium condition” andopening thepreviously “door opening or theopening opening thebetween “initial equilibrium condition” and the “door start” or thestart” “required angle of 75◦ ” good agreement with the numerical predictions, shown in Figure 9“required (abscissa distance angle of 75°” between the points, “initial respectively). equilibrium condition” and the “door opening start” or the “required opening points, respectively). angle of 75°” points, respectively).

Figure Springsactivation activationvs. vs.time: time: activation activation signal displacement Figure 16.16. Springs signal (red), (red),temperature temperature(blue), (blue),and and displacement (black). Figure 16. Springs activation vs. time: activation signal (red), temperature (blue), and displacement (black). (black).

Figure 17. Five activation cycles: right spring contraction (black) vs. door rotation (blue) vs. time. Figure Five activation cycles:right rightspring springcontraction contraction (black) vs. Figure 17.17. Five activation cycles: vs. door doorrotation rotation(blue) (blue)vs. vs.time. time.


17 of 22 17 of 22

Figure Fiveactivation activationcycles: cycles:door dooropening openingand andclosing closing(red (redand andblue bluearrows). arrows). Figure18. 2. Five

A summary and a comparison of the main parameters (numerical and experimental) is provided Table4.4.From From those it resulted good agreement was for theparameters: kinematic in Table those data,data, it resulted that a that goodaagreement was found for found the kinematic parameters: door opening angle vs. 76°), spring total contraction (12 vs. 10.5and mm), and partial spring door opening angle (75◦ vs. 76◦(75° ), spring total contraction (12 vs. 10.5 mm), spring partial contraction until the opening start (4 vs. 3.7 mm). These deviations, due to gaps and free play contraction until the opening start (4 vs. 3.7 mm). These deviations, due to gaps and free play angles angles among the components, system components, then be easily explained could be recovered in a among the system can thencan be easily explained and couldand be recovered in a controlled controlled realization realization process. process. Table 4. 4. Numerical predictions vs experimental experimental outcomes. outcomes.

Type of Extimate Type of Extimate Numerical prediction Numerical Experimental prediction outcome Delta

Door Max Spring Total Spring Contraction Needed forforthe Spring Contraction Energy Energy Needed Door Max Spring Total Opening Angle Contraction before Door Activation of the before Door the Activation Two of Opening Angle (◦ ) Contraction (mm) Opening (mm) theSprings Two Springs (°) (mm) Opening (mm) (J) (J) 75.0

75.0 76.0 *** 1.0

12.0 *

12.010.5 * *** −1.5

4.0 *

4.0 *3.7 *** 0.3

11.8 **

11.8 ** 30.0 **** −18.2

Experimental 76.0 ***by computing 10.5 *** 3.7 *** to move from the pre-load 30.0 ****to the * Quoted in Figure 5. ** Estimated the mechanical work, occurring outcome full activated condition (intersections of the blue curve with the black and the red lines, respectively—Figure 9, 1.0 −1.5 andDelta assuming a typical SMA mechanical efficiency of 0.01 [40]. *** Averaged0.3 values from cycles, plotted−18.2 in Figure 18. **** Estimated value, assuming that the power was constantly supplied for work, 5 s: 3 Woccurring × 5 s × 2 to springs. * Quoted in Figure 5. ** Estimated by computing the mechanical move from the pre-load to the full activated condition (intersections of the blue curve with the black and the red lines, On the contrary, a larger deviationa typical was observed on the energy estimate per *** cycle, necessary respectively—Figure 9, and assuming SMA mechanical efficiency of 0.01 [40]. Averaged to drive thefrom SMA springs (11.8 vs. 30.0 J). The numerical was byconstantly dividing the values cycles, plotted in Figure 18. **** Estimated value,prediction assuming that theobtained power was supplied for 5for s: 3stretching W × 5 s × 2 the springs. mechanical work antagonistic element (geometrically represented by the area under

the blue line in Figure 5, between the pre-load and the fully activated conditions) by the characteristic On the contrary, a larger deviation was observed on the energy estimate per cycle, necessary to efficiency of 1% [40]. The experimental value was instead derived by multiplying the fed power by the drive the SMA springs (11.8 vs. 30.0 J). The numerical prediction was obtained by dividing the dispensing time (5 s). Different factors may concur to this deviation: first, the imperfections in this mechanical work for stretching the antagonistic element (geometrically represented by the area under preliminary realization; second, the friction among the different parts; third, the generic dispersions the blue line in Figure 5, between the pre-load and the fully activated conditions) by the characteristic (electrical and mechanical) of the complete system with the surrounding environment. Substantial efficiency of 1% [40]. The experimental value was instead derived by multiplying the fed power by improvements are then expected when more sophisticated prototypes will be released. the dispensing time (5 s). Different factors may concur to this deviation: first, the imperfections in this preliminary realization; second, the friction among the different parts; third, the generic dispersions 6. Performance vs. Specifications (electrical and mechanical) of the complete system with the surrounding environment. Substantial The work are presented herein when is completed by performing a synthetic improvements then expected more sophisticated prototypes will comparison be released. between the specified (Section 2) and the predicted and demonstrated performance. Data was organized in Table 5. It was aimed at giving a synoptic view of the results for the convenience of the researchers and the


18 of 22

readers. It highlights possible future improvements and deals with the possible drawbacks of the investigated concept. Table 5. Performance vs. specifications. Specifications-Performance

System Size (mm)

Sensor Size (mm)

System Weight (No Sensors) (g)

Power Supply (W)

Requirements Achieved performance Delta Improvement/worsening (↑ or ↓)

≤100 × 30 × 54 99 × 30 × 40 1 × 0 × 14 ↑

≥10 × 2 × 4 = 80 mm3 10 × 10 × 4 = 400 mm3 320 mm3 ↑

≤60 28 32 ↑

≤15 6 9 ↑

Finally, to have an idea of the performance of the SMA actuator with respect to a conventional system used for similar applications (deployment of landing gear of aircraft models, door opening, etc.) [44] a comparison was organized in Table 6. Table 6. Performance comparison with a conventional actuator. Type of Actuator

Weight (g)

Volume (cm3 )

Power (W)

Number of Sensors Potentially Simultaneously Releasable **

Conventional servo-actuator SMA actuator *

39.5 1.0

55.4 1.2

1.5 3.0

10.0 5.0

* Comprehensive of both the springs. ** Computed as the ratio between the power supply available (15 W, Table 5, first row) and the power consumed per cycle.

7. Conclusions and Further Steps In the work at hand the conceptual design and validation of an SMA-based sensor release system for a small UAS was presented. The device was made of a kinematic chain, converting the translational movement of the SMA actuator into the rotation of the fuselage door, allowing the release of sensors and assuring the synchronization among the different movements. After a preliminary schematic was defined, a suitable theoretical model was prepared to link the main system (allowable spring elongation, spring contraction, and so on) to the characteristic SMA parameters (activation temperatures). The model allowed sizing the different components. Its schematic permitted evaluating the adequacy of the synchronization level for properly positioning the sensor onto the expulsion path and timely opening of the panel. The antagonistic device assured the motion reversibility so to perform cyclic operations in order to release many sensors. A preliminary prototype was finally realized to validate the design process and verify the developed concept. Functionality tests proved the capability of the system to perform the established work and verified the correlation among the main operational parameters (temperatures, door rotation, etc.). Moving from these results and the coherence with the established requirements, an advanced design was planned. This new phase was conceived to define the constructive details in order to optimize the system performance before initiating a flight test campaign of characterization and validation in operational conditions. The device integration on-board should be faced by outlining the required interfaces to avoid any interference and to assure the adequate power supply and the coordination with the other existing subsystems. The flight tests are envisaged to be conducted during a typical mission, in the cruise phase at a selected altitude. The main scope of the tests is to demonstrate the effective system ability to match the prescribed operational specifications (for instance, release the sensors according to a prescribed sequence) in clean and perturbed conditions with the ultimate objective to identify the viable flight envelope.


19 of 22

Acknowledgments: This work was developed by the Italian Defence Ministry, in the framework of the Military National Research Program (MNRP) and under the contract CIRA-POI-11-0495, related to activities performed within the Small Unmanned Aerial System design and development. Moreover, the authors would like to recall that the DSC diagram of Figure 8b has been provided by Giovanni Cosentino, Innovative Materials Division of CIRA. Author Contributions: Lorenzo Pellone identified the working principle of the system for releasing the sensors and managed the integrated Digital Mock-Up giving specific suggestions for the system layout. Salvatore Ameduri dimensioned the SMA based sensor release system and organized the experimental functionality campaign. Moving from the UAS needs, Nunzia Favaloro produced the point specifications for the deploying system and provided benchmark info (conventional actuation system). Antonio Concilio provided hints on SMA applications as derived from the related studies and the design approach. Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations a A A AC Af As b B B BC C C CAD CoG c.g. d di D DC DoD DSC Ea ECSS Em ESS F Fact Fantag h K Mf MNRP Ms NMVOC R Rf Rs rcg RPA

length of the sliding element (Figure 7) opening panel pivot (Figure 7) equilibrium state at temperature T1 (Figure 10) rigid connection between the opening panel and the sliding segment (Figure 7) austenite finish temperature: temperature at which, without applied loads, the transformation of martensite into austenite ends austenite start temperature: temperature at which, without applied loads, the transformation of martensite into austenite starts length of the rigid connection AC connection between the active and the antagonistic springs equilibrium state at temperature T3 (Figure 10) sliding element (Figure 7) hinge connection equilibrium state achieved after the activation of the antagonistic spring (Figure 10) Computer Aided Design center of gravity center of gravity of the opening panel slot length of the sliding element (Figure 7) spring wire diameter winding diameter direct current Department of Defense Differential Scanning Calorimetry Austenite Young modulus European Cooperation Space Standardization Martensite Young modulus Environmental Sensor Spreader applied force (Figure 10) force given by the active spring reaction of the antagonistic spring vertical distance between the points A and C (Figure 7) spring rigidity martensite finish temperature: temperature at which, without applied loads, the transformation of austenite into martensite ends Military National Research Program martensite start temperature: temperature at which, without applied loads, the transformation of austenite into martensite starts Non-Methane Volatile Organic Compounds residual of Equation (2) R-phase finish temperature: temperature at which, without applied loads, the transformation of austenite into R-phase ends R-phase start temperature: temperature at which, without applied loads, the transformation of austenite into R-phase starts distance between the center of gravity of the opening door and its pivot, A Remotely Piloted Aircraft


s s SE SMA SUAS T T1 T2 T3 TRL UAS UAV W xb Y ε

20 of 22

linear displacement of the active and antagonistic springs dummy variable for s, used in the definite integral of Equation (5) Super Elastic Shape Memory Alloy Small Unmanned Aerial System temperature initial temperature (Figure 10) intermediate temperature (Figure 10) final warmed temperature (Figure 10) Technology Readiness Level Unmanned Aerial Systems Unmanned Aerial Vehicle Weight of the opening door horizontal displacement of the point B (Equation (3)) Young modulus strain (Figure 10) angle between segment AC and the horizontal line during the opening of the panel (Figure 7) dummy variable for θ, used in the definite integral of Equation (5) initial angle between segment AC and the horizontal line (Figure 7) Poisson modulus martensite fraction stress (Figure 10) angle between the sliding element, BC, and the horizontal line (Figure 7)

θ θ θ0 ν ξ σ ψ

References 1. 2. 3.

4. 5.

6.

7. 8. 9. 10. 11. 12. 13. 14.

Cai, G.; Chen, B.M.; Lee, T.H. An overview on development of miniature unmanned rotorcraft systems. Front. Electr. Electron. Eng. China 2010, 5, 1–14. [CrossRef] Nagy, A.; Rohacs, J. Unmanned Measurement Platform for Paragliders. In Proceedings of the 28th International Congress of the Aeronautical Sciences, Brisbane, Australia, 23–28 September 2012. Eschmann, C.; Kuo, C.-M.; Kuo, C.-H.; Boller, C. Unmanned Aircraft Systems for Remote Building Inspection and Monitoring. In Proceedings of the 6th European Workshop on Structural Health Monitoring, Dresden, Germany, 3–6 July 2012. Jacob, J.; Smith, S.; Cadogan, D.; Scarborough, S. Expanding the Small UAV Design Space with Inflatable Wings; Society of Autmotive Engineers Inc.: Warrendale, PA, USA, 2007. Carpenter, B.; Lyons, J. Lightweight Flexible Solar Array Experiment Summary, NASA Report. 2001. Available online: http://www.eoc.csiro.au/hswww/oz_pi/techforum/Validation_Reports/LFSA.pdf (accessed on 28 November 2016). Petry, G.; Behr, R. The Parafoil Technology Demonstration Project (PTD) lesson learned and Future Visions. In Proceedings of the 15th CEAS/AIAA Aerodynamic Decelerator Systems Technology Conference, Toulouse, France, 9–11 June 1999. Yanushevsky, R. Guidance of Unmanned Aerial Vehicles; CRC Press: Boca Raton, FL, USA, 2011. Cai, G.; Dias, J.; Seneviratne, L. A Survey of Small-Scale Unmanned Aerial Vehicles: Recent Advances and Future Development Trends. Unmanned Syst. 2014, 2, 1–25. [CrossRef] Gupta, A.K.; Arora, S.K. Industrial Automation and Robotics; LAXMI Publications (P) Ltd.: New Delhi, India, 2007. Tai, K.; El-Sayed, A.-R.; Shahriari, M.; Biglarbegian, M.; Mahmud, S. State of the Art Robotic Grippers and Applications. Robotics 2016, 5, 11. [CrossRef] Lecce, L.; Concilio, A. Shape Memory Alloy Engineering: For Aerospace, Structural and Biomedical Applications; Butterworth-Einemann, Elsevier Science: Oxford, UK, 2014. Liang, C.; Rogers, C.A. A multi-dimensional constitutive model for shape-memory alloys. J. Eng. Math 1992, 26, 429–443. [CrossRef] Huang, M.S.; Brinson, L.C. A multivariant model for single crystal shape memory alloy behavior. J. Mech. Phys. Solids 1998, 16, 1379–1409. [CrossRef] Patoor, E.; Lagoudas, D.C.; Entchev, P.; Brinson, L.C.; Gao, X. Shape-memory alloys, part I: General properties and modeling of single crystal. Mech. Mater. 2006, 38, 391–429. [CrossRef]


15.

16. 17.

18. 19. 20. 21. 22. 23. 24.

25. 26. 27. 28. 29. 30.

31. 32. 33. 34.

35. 36. 37. 38.

21 of 22

De Aguiar, R.A.A.; de Castro Leão Neto, W.C.; Savi, M.A.; Calas Lopes Pacheco, P.M. Shape Memory Alloy Helical Springs Performance: Modeling and Experimental Analysis. Mater. Sci. Forum 2013, 758, 147–156. [CrossRef] Yates, S.J.; Kalamkarov, A.L. Experimental Study of Helical Shape Memory Alloy Actuators: Effects of Design and Operating Parameters on Thermal Transients and Stroke. Metals 2013, 3, 123–149. [CrossRef] Villa, T.; Gonzalez, F.; Miljievic, B.; Ristovski, Z.D.; Morawska, L. An Overview of Small Unmanned Aerial Vehicles for Air Quality Measurements: Present Applications and Future Prospectives. Sensors 2016, 16, 1072. [CrossRef] [PubMed] Adam Savage’s tested. Available online: http://www.tested.com/tech/524313-how-get-hobby-rc-testingwise-stabilization/ (accessed on 10 June 2016). RC Group. Available online: http://www.rcgroups.com/forums/showthread.php?t=2260078 (accessed on 20 March 2016). Thing Novation. Available online: http://thingnovation.com/aluminium-robot-clamp-gripper-roboticarduino-garra-brazo-arm-pinza-2-dof.html?___store=en (accessed on 6 February 2016). Thingiverse. Available online: http://www.thingiverse.com/thing:14190 (accessed on 2 July 2016). Robotshop. Available online: http://www.robotshop.com/en/actobotics-horizontal-standard-gripper-kit-a. html (accessed on 13 September 2016). Pollard, E.L.; Jenkins, C.H.M. Shape Memory Alloy Deployment of Membrane Mirrors for Spaceborne Telescopes. J. Spacecr. Rocket. 2007, 44, 1. [CrossRef] Jacobs, S.; Coconnier, C.; Di Maio, D.; Scarpa, F.; Toso, M.; Martinez, J. Deployable auxetic shape memory alloy cellular antenna demonstrator: Design, manufacturing and modal testing. Smart Mater. Struct. 2012, 21, 1–12. [CrossRef] Calkins, F.T.; Mabe, J.H. Shape Memory Alloy Based Morphing Aerostructures. J. Mech. Des. 2010, 132, 111012. [CrossRef] Sofla, A.Y.N.; Elzey, D.M.; Wadley, H.N.G. Shape morphing hinged truss structures. Smart Mater. Struct. 2009, 18, 1–8. [CrossRef] Unsworth, J.D.; Waram, T.C. Shape Memory Alloy Tubular Deployment System. U.S. Patent 5,846,247, 8 December 1998. Dayananda, G.N.; Varughese, B.; Rao, M.S. Shape Memory Alloy Based Smart Landing Gear for an Airship. J. Aircr. 2007, 44, 1469–1477. [CrossRef] Unmanned Aerial Vehicles Reliability Study. Available online: http://www.uadrones.net/military/ research/acrobat/0302.pdf (accessed on 6 December 2003). Abdullah, E.J.; Bil, C.; Watkinsm, S. Testing of Adaptive Airfoil for UAV using Shape Memory Alloy Actuators. In Proceedings of the 27th International Congress of the Aeronautical Sciences, Nice, France, 19–24 September 2010. Ameduri, S.; Concilio, A.; Favaloro, N.; Pellone, L. A Shape Memory Alloy Application for Compact Unmanned Aerial Vehicles. Aerospace 2016, 3, 16. [CrossRef] Madhani, A.J. Robot Hand with Human-Like Fingers. EP Patent 2,239,106 B1, 14 March 2012. Chaitanya, S.K.; Dhanalakshmi, K. Design and investigation of a shape memory alloy actuated gripper. Smart Struct. Syst. 2014, 14, 4541–4558. Clausi, D.; Gradin, H.; Braun, S.; Peirs, J.; Stemme, G.; Reynaerts, D.; van der Wijngaart, W. Robust actuation of silicon MEMS using SMA wires integrated at wafer-level by nickel electroplating. Sens. Actuators A 2013, 189, 108–116. [CrossRef] Bundhoo, V.; Haslam, E.; Birch, B.; Park, E.J. A shape memory alloy-based tendon-driven actuation system for biomimetic artificial fingers, part I: Design and evaluation. Robotica 2009, 27, 131–146. [CrossRef] Prapaipittayakhun, K.; Srituravanich, W.; Pimpin, A. Shape Memory Alloy Micro-Actuator for Handling of Head Gimbal Assembly. Aijstpme 2011, 4, 1–8. Machado, L.G.; Savi, M.A. Medical applications of shape memory alloys. Braz. J. Med. Biol. Res. 2003, 36, 683–691. [CrossRef] [PubMed] Pecora, R.; Iannuzzo, G.; Riccio, M.; Russo, S.; Calvi, E.; Lecce, L.; Barbarino, S.; Concilio, A.; Ameduri, S. Actuator Device Based on a Shape Memory Alloy, and a Wing Flap Assembly Fitted with Such an Actuator Device. U.S. Patent 8,348,201 B2, 8 January 2013.


39. 40. 41. 42.

43.

44.

22 of 22

Jani, J.M.; Leary, M.; Subic, A. Shape Memory Alloys in Automotive Applications. Appl. Mech. Mater. 2014, 663, 248–253. [CrossRef] Huber, J.E.; Fleck, N.A.; Ashby, M.F. The selection of mechanical actuators based on performance indices. Proc. R. Soc. Lond. A 1997, 453, 2185–2205. [CrossRef] Barooah, P. System and Method for Controlling Shape Memory Alloy Actuators. U.S. Patent 6,543,224 B1, 8 April 2003. Ghomshei, M.M.; Khajepour, A.; Tabandeh, N.; Behdinan, K. Finite Element Modeling of Shape Memory Alloy Composite Actuators: Theory and Experiment. J. Intell. Mater. Syst. Struct. 2001, 12, 761–773. [CrossRef] Serry, M.Y.; Moussa, W.A.; Raboud, D.W. Finite-element modeling of shape memory alloy components in smart structures, part II: Application on shape-memory-alloy-embedded smart composite for self-damage control. In Proceedings of the International Conference on MEMS, NANO and Smart Systems, Banff, AB, Canada, 23 July 2003. Hobby Modellismo. Available online: http://www.hobbymodellismo.com/it/carrello-retrattile-elettricolaterale-dy1023 (accessed on 6 July 2016). © 2017 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/).

SMA-Based System for Environmental Sensors Released from ... - MDPI

SMA-Based System for Environmental Sensors Released from ... - MDPI

Suggest Documents

Pitot Sensors System for Aircraft Velocity - MDPI

Sensors for Entertainment - MDPI

Overview of Sensors and Needs for Environmental Monitoring - MDPI

An Environmental Monitoring System for Managing ... - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI

sensors - MDPI