Composites Part B 78 (2015) 507e514
Contents lists available at ScienceDirect
Composites Part B journal homepage: www.elsevier.com/locate/compositesb
Smart soft composite actuator with shape retention capability using embedded fusible alloy structures Wei Wang a, Hugo Rodrigue a, Sung-Hoon Ahn a, b, * a b
Department of Mechanical and Aerospace Engineering, Seoul National University, Seoul 151-742, Republic of Korea Institute of Advanced Machinery and Design, Seoul National University, Seoul 151-742, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 February 2015 Received in revised form 25 March 2015 Accepted 3 April 2015 Available online 11 April 2015
This work presents a new kind of shape memory alloy (SMA) based composite actuators that can retain its shape in multiple configurations without continuous energy consumption by changing locally between a high-stiffness and a low-stiffness state. This was accomplished by embedding fusible alloy (FA) material, Ni-chrome (NieCr) wires and SMA wires in a smart soft composite (SSC) structure. The soft morphing capability of SMA-based SSC structures allows the actuator to produce a smooth continuous deformation. The stiffness variation of the actuator was accomplished by melting the embedded FA structures using NieCr wires embedded in the FA structure. First, the design and manufacturing method of the actuator are described. Then, the stiffness of the structure in the low and high-stiffness states of the actuator were measured for different applied currents and heating durations of the FA structure and results show that the highest stiffness of the actuator is more than eight times that of its lowest stiffness. The different shape retention capability of the actuator were tested using actuators with one or two segments and these were compared with a numerical model. © 2015 Elsevier Ltd. All rights reserved.
Keywords: A. Polymer-matrix composites (PMCs) A. Smart materials B. Mechanical properties Shape retention
1. Introduction Certain biological muscular systems such as the trunk of the elephant and the arm of the octopus are able to locally switch between a low-stiffness state when interacting with the external environment and a high-stiffness state when manipulating objects. This principle has been an inspiration for new biomimetic structures capable of changing their stiffness locally or globally, deforming under internal or external forces and retaining their deformed shape such that they can withstand external loads and stimuli. These materials and structures capable of modulating their stiffness have become an integral element of various applications ranging from medical science to mechanical engineering. Structures with variable stiffness were used in tools to conduct colonoscopies to improve the reliability and avoid tissue damage [1,2]. Light-weight morphing structures with variable stiffness were used in an adaptive building skin capable of modifying energy flows, reducing noise and preventing fires [3]. Structures capable of changing their stiffness under actuation can be applied to shape
* Corresponding author. Department of Mechanical and Aerospace Engineering, Seoul National University, Seoul 151-742, Republic of Korea. Fax: þ82 2 8889073. E-mail address:
[email protected] (S.-H. Ahn). http://dx.doi.org/10.1016/j.compositesb.2015.04.007 1359-8368/© 2015 Elsevier Ltd. All rights reserved.
reconfigurable structures such as morphing wings for micro air vehicles [4,5]. Furthermore, soft robots have become a popular field of research [6] and their capabilities could be enhanced by implementing sub-structures that could change between being soft and hard-bodied based on the different tasks they need to accomplish. This kind of structure has been implemented by varying the temperature of materials with low melting temperatures such as thermoplastics [7] and fusible alloys [8] embedded in a rigid or soft matrix. However, the structures presented in these studies are not capable of deforming themselves into complex user-defined shapes. Mechanical structures comprised of components such as motors, linkages and gears can be used to create structure capable of retaining specific shapes, but their use increases the system's complexity and they are not capable of producing complex and continuous deformations. Pneumatic soft actuators can generate complex deformations and retain their deformed configuration by controlling the internal air pressure of the deforming structure. However, because they require an air compressor and air valves, it is difficult for these actuators to be built into an independent system at small scales [9,10]. Adaptive composite structures with multi-stable configurations can be used as morphing structures with good shape retention capabilities and can withstand adequate loads, but they possess limited deformation capabilities and can
508
W. Wang et al. / Composites Part B 78 (2015) 507e514
only retain specific combinations of shapes [11,12]. EAP actuators are an attractive option for robotic applications [13,14], but the relatively small actuation force, short lifetime, high response times or high driving electric fields of these EAP materials limits their applications and there is no research on EAP actuators with shape retention. Shape memory polymers (SMPs) with a thermotemporal shape memory cycle can be deformed at high temperatures and then fixed upon cooling to retain a bent or twisted shape [15]. However, SMPs have a longer response time and produce smaller actuation stresses in comparison with shape memory alloys (SMAs) [16]. Although SMA elements have a small working strain, soft composite actuators with embedded SMA elements were developed that are capable of large, continuous and complex deformations. Smart soft composite (SSC) actuators consisting of a low-stiffness matrix with embedded SMA elements [17]. Following research has developed SSC actuators with bending, twisting and bending-twisting motions [18e20]. However, since SMA elements need to maintain a certain temperature to remain in the deformed shape, this type of structure needs constant energy input to maintain the desired deformation. Because of this and the low energy efficiency, long-term actuation of SMA elements is not a viable option due to the high energy requirements [21]. An SMAbased woven-type composite structure has been studied in Ref. [22] which is capable of retaining its deformed shape after bending, but this structure was capable of a limited deformation and the loading capability of the structure was not tested. In this research, an SMA-based SSC actuator with shape retention capability is presented. The transformation from a highstiffness structure to a low-stiffness one was accomplished by melting the embedded fusible alloy (FA) structures using Nichrome (NieCr) wires. Afterwards, SMA wires are used to achieve a large continuous deformation of the structure and maintain the shape of the structure while the FA structures solidifies. Then, current input to the SMA wires can be stopped and the structure can maintain its shape without requiring constant energy input. First, the fabrication method of the actuator is described, then the flexural stress of the actuator is tested for different operating conditions. Afterwards, the shape retention capability of the actuator was tested for an actuator with a single segment and for an actuator with two segments. Finally, a model is proposed to predict the shape of the actuator's final retained configurations and is compared with the obtained results. 2. Materials and methods 2.1. Materials The material used for the matrix of the actuator is Polydimethylsiloxane (PDMS, Dow Corning Sylgard 184). This material was selected due to its high flexibility, low thermal conductivity and high thermal stability [23]. Its main properties are listed in Table 1. The SMA wires are Flexinol wires (55 wt% Ni, 45 wt% Ti, Dynalloy, US) and their properties are listed in Table 2 [24]. There are many materials with low melting points including
Table 1 Main properties of PDMS (sylgard 184).
Table 2 Material properties of SMA (flexinol). Parameter
Value
Martensitic Young's modulus Austenitic Young's modulus Martensitic start temperature Martensitic finish temperature Austenite start temperature Austenite finish temperature Wire diameter Resistance per meter Initial strain
EMar ¼ 28 GPa EAus ¼ 75 GPa Ms ¼ 52 C Mf ¼ 42 C As ¼ 68 C Af ¼ 78 C 0.152 mm 55 U ε0 ¼ 5%
thermoplastics, such as polycaprolactone (PCL), and fusible alloys (FA), such as Field's metal, Lipowitz's alloy or Wood's metal. In this work, the FA Field's metal (RotoMetals, Inc) with a melting point of 62 C was selected. This FA is an eutectic alloy of bismuth, indium, and tin (32.5 wt% Bi, 51 wt% In, 16.5 wt% Sn) and was selected due to its high thermal conductivity, low viscosity in the melted state, high-stiffness in the solid state, its non-toxicity, its melting point falls within the temperature range of the PDMS matrix and is lower than the SMA Austenite starting temperature. In order to melt the FA structure, Joule heating of a NieCr (80 wt% Ni, 20 wt% Cr) wire is used [25]. The NieCr wire has a diameter of 0.15 mm and is covered with a polyimide (PI) tube (D. Soar Green, China), which has a good thermal conductivity and are electrically non-conductive, to prevent the current from going through the fusible alloy. The dimensions and relevant material properties of the PI tubes are listed in Table 3.
2.2. Design and fabrication The proposed actuator has a rectangular shape where two SMA wires, referred to as SMA-1 and SMA-2, are embedded in the PDMS matrix along the Y axis close to the upper and lower surfaces, as shown in Fig. 1. Two symmetrical FA structures are placed in parallel to the SMA wires in the middle of the matrix. The relative position of all the components is shown in the schematic of crosssection AeA of the actuator in Fig. 1. The actuation process to change from one fixed shape to another is shown in Fig. 2 with the actuation sequence for the NieCr wires and the SMA wires shown in Fig. 2a. During this process, the embedded FA structures are melted by applying current to the NieCr wires from time t0 until time t1. Then, the current to the NieCr wires is switched off and actuating either SMA-1 or SMA-2 from time t1 to time t2 will result in either an upward or downward bending deformation depending on which SMA wire is actuated. Then, current continues to be applied to the actuated SMA wire until time t3 to allow the melted FA structures to solidify again. Fig. 2b to e shows the state of the actuator from time t0 to time t3. The SMA wires were clamped with connectors to connect with the conductive wire and to prevent relative sliding between the SMA wires and the matrix [26]. The FA structure was built by
Table 3 Material properties of PI tube (D. soar green).
Parameter
Value
Parameter
Value
Temperature range Specific gravity Heat cure PDMS Young modulus Thermal conductivity Volume resistivity
45 to 200 C 1.03 @ 25 C 12 h @ 50 C EPDMS ¼ 1.8 MPa @ 25 C 0.27 W/m K 2.9 1014 U-cm
Inner diameter Thickness Density Temperature limit Dielectric strength Heat conductivity coefficient
0.18 mm 0.015 mm 1.41 103 kg/m3 380 C >1.18 1010 kV/m 35.0 105 C/cm
W. Wang et al. / Composites Part B 78 (2015) 507e514
509
into the mold. Fourth, a flat acrylic plate is positioned on top of the mold to get an even surface. Fifth, the assembly is placed in an oven and cured at a temperature of 50 C for 12 h. After the curing process, the specimen is taken out of the oven and the ABS mold is removed. The complete structure of the actuators with one SMA wire segment (upper) and two SMA wire segments (below) with overall actuator dimensions of 100 20 3 mm (length width thickness) are shown in Fig. 3e.
3. Results and discussion 3.1. Structural variable stiffness characteristics Fig. 1. The actuator configuration and its components.
casting using a three dimensional printed mold made from acrylonitrile butadiene styrene (ABS). The ABS mold is manufactured by fused deposition modeling (FDM) (Dimension 768 SST, Stratasys, USA). The NieCr wire covered with a PI tube is positioned in the mold and fixed mechanically using screws located on an external jig. The FA is melted by submerging it in water at 90 C, drawn into a syringe and injected into the mold as shown in Fig. 3a. After injecting the FA into the mold and allowing the assembly to cool for 20 s, the FA structure is then removed from the mold and the structure shown in Fig. 3c is obtained. An enlarged view of the tip of the actuator is shown in Fig. 3d showing the positions of the NieCr wire and of the PI tube within the FA structure. The NieCr wires can be heated segmentally by removing a portion of the Polyimide tube around the NieCr wire and by connecting the current to one end of the NieCr wire and the other end to the FA structure. The schematics of the FA structure with a single segment and with two segments are shown in Fig. 3b where a current passing from A to C in the structure with two segments will melt the left segment of FA structure and an electric current passing from B to C will melt the right segment of the structure. By using the FA structure with two segments with a connector positioned at the halfway point of the SMA wire, it is possible to actuate only a portion of the actuator. A second ABS mold made by the same FDM machine as the FA structure's mold is used to manufacture the actuator. First, the two FA structures are positioned in the mold and mechanically fastened using bolts located on an external jig. Second, the two SMA wires with connectors are positioned in the mold with a pre-strain of 5% and mechanically fastened using bolts located on the same jig. Third, a PDMS solution with a weight ratio of 10:1 monomer to hardener is mixed, degassed in a vacuum casting pump and poured
The variable stiffness property of the actuator stems from the difference in stiffness between the solidified and the melted FA structure. In order to quantify this difference in stiffness, the flexural property of the actuator was measured using a three-point loading test according to the ASTM D 790-03 standard test method. The test was done using a span-to-depth ratio of 16:1 on a tensile testing machine (5948 MicroTester, Instron, US) with a 3point flexure fixture (2810-400, Instron, US). To change the overall stiffness of the actuator, a series of electrical current of 0.2 A, 0.4 A, 0.5 A, 0.6 A, 0.8 A and 1.0 A were applied to the NieCr wire to heat the embedded FA structures for a fixed time period of 90 s. By applying current to the NieCr wire, the FA structure within the actuator heats up gradually and turns into the liquid phase. The stiffness of the structure decreases in parallel with this change in phase of the FA structure. It was determined through experiments that the structure neither yields nor breaks before the 5% strain limit of the testing method, therefore the flexural stress (sf) which can be calculated as in (1) is taken at the 5% strain limit.
. sf ¼ 3PL 2bd2
(1)
where P is the load applied to middle point of the specimen, L is the supporting span of the three point bending test, and b and d are the width and thickness of the tested actuator specimen. The results are shown in Fig. 4, from which it can be seen that the decrease in flexural stress undergoes three distinct stages depending on the magnitude of the applied current. In the first stage, for a current smaller than 0.2 A, the flexural stress decreases very little due to insufficient heating of the FA structures to begin the melting process. In the second stage, for currents between 0.2 A and 0.8 A, the flexural stress decreases when the current is
Fig. 2. The shape retention process. (a) The actuation sequence. (b) to (e) are the different states of the actuator from time t0 to time t3.
510
W. Wang et al. / Composites Part B 78 (2015) 507e514
Fig. 3. Fabrication process of the actuator. (a) FA structure fabrication setup. (b) to (d) Fabricated FA structure. (e) Fabricated shape retention actuator.
increased, which shows that the FA structures change from an incompletely melted state to a completely melted state between these currents. For currents of 0.8 A or higher, it can be seen that the flexural stress is constant, which shows that the FA structure is completely melted. In order to melt the FA structure faster, a higher current can be used to accelerate the increase in temperature of the structure. However, at both ends of the matrix, the NieCr wire and the PI tube are in direct contract with the PDMS matrix, which has an upper temperature limit of 200 C and the upper temperature limit of the PI tube itself is 380 C. The temperature of the NieCr wires should remain within the temperature limit of the PDMS and of the PI tube. In addition, a higher current will generate more redundant heat. To avoid these problems while still having rapid heating of the FA structures, a current of 1.0 A was selected for the NieCr wires. Next, an experiment was conducted for different heating periods of the FA structure with an applied current to the NieCr wires of 1.0 A. Tests were conducted for heating times ranging from 0 s to 60 s with time increments of 10 s to determine how long the NieCr wire should be heated before actuating the SMA wires. A cooling time of 10 min was given between each test. The results are shown in Fig. 5, from which it can be seen that the flexural stress becomes nearly constant for heating times greater than 40 s. This indicates that the embedded FA structures are completely in the melted phase. From the first two experiments, it can be seen that the flexural stress of the non-heated sample (3.34 MPa) is eight times that of the sample (0.41 MPa) after complete melting of the FA structures.
Fig. 4. Flexural stress of the actuator at 5% strain based on the applied current.
The third set of experiments were conducted to test the cooling time of the actuator at a room temperature of 23 C after continuous heating of the FA structures for 50 s at a current of 1.0 A, which was shown to fully melt the FA structure in the previous experiment. Experiments were conducted in increments of 30 s up to 300 s and in increments of 100 s from 300 s to 600 s. It can be seen from the experimental results shown in Fig. 6 that the flexural stress of the actuator increases significantly with the cooling time until 240 s and is nearly constant afterwards. This indicates that it takes approximately 240 s for the structure to cool down completely. 3.2. Structural shape retention characteristics 3.2.1. Shape retention sequence The actuating current for the SMA wires and the required current for heating of the NieCr wires embedded in the matrix with diameters of 0.152 mm were determined to be 0.55 A and 1.0 A, respectively. The time t1 was determined to be 50 s in previous experiments and the time t2 corresponds to the heating time plus the cooling time, which corresponds to 290 s based on the results of the previous experiments. Fig. 7a describes the time sequence of the actuator, and Fig. 7b to d shows the position of the actuator for bending in the left direction with different cooling times of t0 ¼ 0 s,
Fig. 5. Flexural stress of the actuator at 5% strain based on the heating time.
W. Wang et al. / Composites Part B 78 (2015) 507e514
511
Fig. 6. Flexural stress of the actuator after heating for 50 s for different cooling times.
t1 ¼ 50 s, t2 ¼ 290 s and t3 ¼ 1800 s. In Fig. 7e, the positions of the actuator at times t2 and t3 are juxtaposed for comparison. From this comparison, it can be seen that there is a slight recovery after stopping the actuation of the SMA wire, which is mainly be due to removing the bending force from the actuated SMA wire and from the recovery force from the unactuated SMA wire. 3.2.2. Shape retention of two-segmented actuator Section 3.2.1 described the shape retention sequence of a shape retention actuator with a single segment, which is capable of maintaining three configurations: straight, left-bending and rightbending shapes. By building a segmented actuator, it is possible to obtain an actuator capable of deforming into a greater amount of shapes of increased complexity and of retaining those shapes. Each segment is capable of having three configurations: straight, leftbending and right-bending. Thus, depending on which segments of the FA structures is melted and which portion of the SMA wires is actuated, this actuator is capable of achieving 3n final configurations where n corresponds to the number of segments. In this section, an actuator with two segments as described in section 2.2 was tested. The configuration of this actuator was shown in Fig. 3e (bottom). The nine configurations of an actuator with two segments were obtained experimentally using a heating time of 50 s with a current of 1.0 A and a cooling time of 240 s and are shown in Fig. 8 where the two segments of the structures are shown in yellow and red (in the web version).
Fig. 8. Nine different configurations of the two-segmented actuator, (a) to (i).
3.2.3. Energy comparison In comparison with an actuator without shape retention capability, one of the main advantages of an actuator with shape retention capability is energy savings for long-term deformation of a structure since it does not continuously need an applied current to maintain its deformed shape. To attain its deformed shape, the actuator must first melt the embedded FA structure, then deform itself and maintain its shape until the FA structure solidifies again. After this, no energy input is required. A similar actuator without shape retention capability would need a constant and continuous energy input. The energy consumption of SSC actuators with and without the proposed shape retention mechanism are calculated and shown in Fig. 9. From these results, it can be seen that until approximate 425 s the energy consumption of the actuator with shape retention capability is bigger than that of the actuator without shape retention capability due to the energy required for heating the NieCr wire being higher than the energy required for heating the SMA wire during the same period of time. After 290 s, the actuator with shape retention capabilities stops consuming energy and at 425 s the total consumption of both actuators is equal. After 425 s, the actuator without shape retention capability will continue consuming approximate 1.66 J/s while the actuator with shape retention capability will not consume any energy, it thus becomes beneficial to use an actuator with shape retention
Fig. 7. Shape retention sequence diagram. (a) Current sequence for actuating and cooling. (b) Actuator configuration before actuation of SMA wire. (c) to (d) The left-bending shape at time t2 and t3, and (e) juxtaposed.
512
W. Wang et al. / Composites Part B 78 (2015) 507e514
Fig. 11. Schematic of an actuator with a single segment before actuation (a) and after actuation of SMA-1 (b).
Fig. 9. Energy comparison of the actuator with and without shape retention capability.
capabilities if the predicted time in the deformed shape is greater than 425 s.
where ε0 is the SMA initial strain, I is the moment of inertia of the cross-section of the actuator and ASMA is the cross section area of the SMA wires. Using the value calculated as in (1), the bending angle q1 of the actuator can be calculated as shown in Eq. (3).
q1 ¼ ðε0 εÞL1 =d
(3)
3.3. Mechanical model 3.3.1. Bending model of actuator structure First, the Young's modulus of the actuator with the melted and solid FA structures was calculated using the experimental data from Section 3.1 and the results are shown in Fig. 10. The Young's modulus of the actuator with the FA structures in the melted state and solid state are EL-Stru ¼ 13.23 MPa and EH-Stru ¼ 274.97 MPa, respectively. The modeling of SMA-based soft composite structures capable of bending deformations has been studied previously [27e30]. The configurations of the actuator before and after actuation, and the notation of the different variables used subsequently are shown in Fig. 11. The instantaneous SMA strain ε induced by resistance to the SMA free thermal contraction for a flexible composite with embedded SMA wires was derived in Ref. [27] and can be calculated as shown in (2) with the assumption that the SMA wires have zero initial stress, a negligible thermo-elastic term, and a fully austenite transformation.
. EAus ASMA d2 þ ELStru I ε ¼ ELStru Iε0
(2)
where L1 is the length of the actuator and d is the distance between the SMA wires and the neutral plane of the actuator. After reaching the maximum bending angle q1 and cooling of the FA structures, the current to SMA-1 is switched off and there is a slight recovery force from the deformed actuator structure that makes the actuator undergo a recovery of angle q2. Considering the Young's modulus of PDMS is much smaller than that of the SMA wires, the main recovery force was considered to stem from the strain recovery of SMA-2. Assuming that the length of actuator structure's neutral place remains unchanged, the strain of SMA-1 is reduced by ε0ε for the deformation shown in Fig. 11b and the strain for SMA-2 is increased by ε0ε. Therefore, the strain of SMA2 in this state is equal to ε0þε0ε ¼ 2ε0ε. Referring to (2), the strain ε0 of SMA-2 induced by resistance to the SMA free contraction can be calculated as in (4).
ε0 ¼ EHStru Ið2ε0 εÞ
. EMar ASMA d2 þ EHStru I
(4)
Then, the recovered bending angle q2 of the actuator can be calculated as in (5).
q2 ¼ ð2ε0 ε ε0 ÞL1 =d
(5)
Thus, the final bending angle q for a one-segmented actuator will be expressed as q ¼ q1 q2. The parameters used in calculations and the results obtained from the model are shown in Table 4. 3.3.2. Multiple configurations The position of the end point is used in order to quantitatively describe the final retained shape of the actuator after cooling of the actuator for the different configurations. The final retained bending angle q of the actuator can be determined theoretically and depends on the configuration of the actuator. For an actuator with a single segment, the position of the actuator's end point O1 with respect to the fixed point O before Table 4 Parameters of actuator components.
Fig. 10. Young's modulus of the actuator with the FA structure in the melted state and in the solid state.
Parameter
Value (unit)
d L1
1 mm 50 mm 57.80 6.42
q1 q2
W. Wang et al. / Composites Part B 78 (2015) 507e514
actuation, as shown in Fig. 11a, can be expressed using the homogeneous transformation matrix as in (6).
2
HOO1
1 ¼ 40 0
0 1 0
3 L1 05 1
(6)
In this matrix, the first two numbers of the last column are the coordinates of the end point (L1,0) in the inertial frame. After actuation, as shown in Fig. 11b, the coordinate of the end-point O01 and the direction axes x01 and y01 fixed on the endpoint have a translation and rotation motion with a counterclockwise bending angle of q with respect to the fixed coordinate of the fixed base point O and of its direction axes x and y. The distance L2 between the deformation tip of the actuator O01 and the fixed base point of the actuator O can be calculated as in (7):
L2 ¼
2L1 ð±qÞ sin 2 q
(7)
where ±q corresponds to the clockwise and anti-clockwise bending angle of the retained deformation shape for one segment. Then, the position of the actuator end point O01 with respect to the fixed point O can be expressed using the homogeneous transformation matrix as in (8):
2
HOO2 ;±q
cosð±qÞ ¼ 4 sinð±qÞ 0
sinð±qÞ cosð±qÞ 0
3 L2 cosð±q=2Þ L2 sinð±q=2Þ 5 1
(8)
(6) and (8) are basic transformation matrixes. Therefore, for an nsegmented actuator, the end point position with respect to the fixed base point can be expressed as in (9) and the first two numbers of the last column of this matrix are the coordinates of the end point with respect to the inertial frame.
HOOn ¼
n Y
HOi1 Oi ;±q
(9)
i¼1
For the two-segmented actuator presented in this study, the nine configurations of the shape retention actuator were recorded using a camera and the coordinates of the middle points and of the end points were measured visually. The experimentally measured data and the predicted values obtained through the model are shown in Fig. 12. In this figure, the nine configurations of the actuator obtained from the modeling are juxtaposed and shown with gray lines. In this experiment, only the in-plane shape retention capability of the actuator was tested. There are some
513
errors between the experimental and numerical results which could be due to imprecisions in the properties of the actuator and the experimental measurement method, but the developed model corresponded well with the experimental data. The PDMS matrix material is much softer than the SMA material, so when actuating SMA-1 to achieve the maximum deformation, the distance from SMA-1 to the actuator's neutral surface should be a little larger than d due to the SMA wire being able to deform slightly the matrix, and the distance from SMA-2 to the actuator neutral surface should be a little smaller than d. However, in the model, the distances were assumed be constant. In addition, the experimental data was obtained using a camera positioned manually, which could lead to some visual measurement errors. 4. Conclusion This study first described an SMA-based variable stiffness actuator which relies on FA structures embedded in a SSC actuator capable of both soft morphing and of retaining multiple shape configurations. By designing an actuator with multiple segments, this actuator can obtain and retain complex shapes. This is achieving by both changing its local structural stiffness and deforming the corresponding segments to deform locally the structure. The time for changing from highest stiffness to lowest stiffness of the actuator was determined as 50 s and that for the inverse process was determined as 240 s. Therefore this method is suitable when there is no strict time requirements for changing the stiffness of structures. The advantage of this composite structure is that the properties of this structure can be tailored to its application by modifying the volume, position, geometry and composition of the embedded low melting material structures. Applications for this type of actuator range from medicine science to transformable or dynamic mechanical structures. Further research will focus on improving the performance of the presented actuator and on improving the capabilities of the actuator by applying its operational principle to structures capable of retaining shapes that have more complex shape profiles, such as in three-dimensions. Also, the proposed structures could be combined with soft sensors to obtain feedback on the current shape of the structure. Acknowledgments This work was supported by the Industrial Strategic technology development program (10049258) funded by the Ministry of Knowledge Economy (MKE, Korea) and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (No. NRF-2010-0029227). The deepest gratitude is expressed to the providence of God. References
Fig. 12. Configurations of modeling and measured coordinates of middle and end points.
[1] Cheng WB, Moser MA, Kanagaratnam S, Zhang WJ. Overview of upcoming advances in colonoscopy. Dig Endosc Off J Jpn Gastroenterol Endosc Soc 2012;24:1e6. [2] Raju GS, Rex DK, Kozarek RA, Ahmed I, Brining D, Pascricha PJ. A novel shapelocking guide for prevention of sigmoid looping during colonoscopy. Gastrointest Endosc 2004;59:416e9. [3] Nguyen QT, Tran P, Ngo TD, Tran PA, Mendis P. Experimental and computational investigations on fire resistance of GFRP composite for building facade. Compos Part B Eng 2014;62:218e29. [4] Han MW, Rodrigue H, Cho SH, Song SH, Chu WS, Ahn SH. Design and performance evaluation of soft morphing car-spoiler. In: Proceedings of ASME 2014 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference. New York; August, 2014. V05AT08A037. [5] Bettini P, Airoldi A, Sala G, Landro LD, Ruzzene M, Spadoni A. Composite chiral structures for morphing airfoils: numerical analyses and development of a manufacturing process. Compos Part B Eng 2010;41(2):133e47.
514
W. Wang et al. / Composites Part B 78 (2015) 507e514
[6] Shepherd RF, Ilievski F, Choi W, Morin SA, Stokes AA, Mazzeo AD, et al. Multigait soft robot. Proc Natl Acad Sci 2011;108:20400e3. [7] McEvoy MA, Correll N. Thermoplastic variable stiffness composites with embedded, networked sensing, actuation, and control. J Compos Mater 2014;0:1e10. [8] Schubert BE, Floreano D. Variable stiffness material based on rigid lowmelting-point-alloy microstructures embedded in soft poly (dimethylsiloxane) (PDMS). RSC Adv 2013;3:24671e9. [9] Martinez RV, Glavan AC, Keplinger C, Oyetibo AI, Whitesides GM. Soft actuators and robots that are resistant to mechanical damage. Adv Funct Mater 2014;24:3003e10. [10] Mosadegh B, Polygerinos P, Keplinger C, Oyetibo AI, Whitesides GM. Pneumatic networks for soft robotics that actuate rapidly. Adv Funct Mater 2014;24:2163e70. [11] Zhang Z, Wu H, He X, Wu H, Bao Y, Chai G. The bistable behaviors of carbonfiber/epoxy anti-symmetric composite shells. Compos Part B Eng 2013;47: 190e9. [12] Lee JG, Ryu J, Lee H, Cho M. Saddle-shaped, bistable morphing panel with shape memory alloy spring actuator. Smart Mater Struct 2014;23:074013. €plin B, Gülch RW. Electroactive polymer (EAP) actua[13] Wallmersperger T, Kro tors as artificial muscles-reality, potential, and challenges. Modelling and Analysis of Chemistry and Electromechanics, vol. PM136; 2004. p. 335e62. [14] Carpi F, Kornbluh R, Sommer-Larsen P, Alici G. Electroactive polymer actuators as artificial muscles: are they ready for bioinspired applications? Bioinspir Biomim 2011;6(4):045006. dy C, Gilormini P, Merckel Y, Re gnier G, Rousseau I. A torsion test [15] Diani J, Fre for the study of the large deformation recovery of shape memory polymers. Polym Test 2011;30:335e41. [16] Guo X, Liu L, Liu Y, Zhou B, Leng J. Constitutive model for a stress-and thermalinduced phase transition in a shape memory polymer. Smart Mater Struct 2014;23:105019. [17] Ahn SH, Lee KT, Kim HJ, Wu R, Kim JS, Song SH. Smart soft composite: an integrated 3D soft morphing structure using bend-twist coupling of anisotropic materials. Int J Precis Eng Manuf 2012;13:1e4.
[18] Wang W, Lee JY, Rodrigue H, Song SH, Chu WS, Ahn SH. Locomotion of inchworm-inspired robot made of smart soft composite. Bioinspir Biomim 2014;9:046006. [19] Kim HJ, Song SH, Ahn SH. A turtle-like swimming robot using a smart soft composite (SSC) structure. Smart Mater Struct 2013;22:014007. [20] Rodrigue H, Wang W, Bhandari B, Han MW, Ahn SH. Cross-shaped twisting structure using SMA-based smart soft composite. Int J Precis Eng Manufacturing Green Technol 2014;1:153e6. [21] Langbein S, Czechowicz A. Adaptive resetting of SMA actuators. J Intell Mater Syst Struct 2012;23:127e34. [22] Wu R, Han MW, Lee GY, Ahn SH. Woven type smart soft composite beam with in-plane shape retention. Smart Mater Struct 2013;22:125007. [23] Schneider F, Fellner T, Wilde J, Wallrabe U. Mechanical properties of silicones for MEMS. J Micromech Microeng 2008;18:065008. [24] Gilbertson RG. Muscle wires project book. California, USA: Mondo-Tronics, Inc; 1994. [25] She Y, Li C, Cleary J, Su HJ. Design and fabrication of a soft robotic hand with embedded actuators and sensors. J Mech Robotics 2015;7:021007. [26] Kim DJ, Kim HA, Chung YS, Choi ES. Pullout resistance of straight NiTi shape memory alloy fibers in cement mortar after cold drawing and heat treatment. Compos Part B Eng 2014;67:588e94. [27] Smith C, Villanueva A, Joshi K, Tadesse Y, Priya S. Working principle of bioinspired shape memory alloy composite actuators. Smart Mater Struct 2011;20:012001. [28] Bodaghi M, Damanpack AR, Aghdam MM, Shakeri M. Active shape/stress control of shape memory alloy laminated beams. Compos Part B Eng 2014;56: 889e99. [29] Dawood M, El-Tahan MW, Zheng B. Bond behavior of superelastic shape memory alloys to carbon fiber reinforced polymer composites. Compos Part B Eng 2015;77:238e47. http://dx.doi.org/10.1016/j.compositesb.2015.03.043. [30] Jia J, Rogers CA. Formulation of a mechanical model for composites with embedded SMA actuators. J Mech Des 1992;114:670e6.