Academia.eduAcademia.edu

Figure from:A Vacuum Powered Soft Textile-Based Clutch

See full PDFdownloadDownload figure
Figure 5. (a) The control unit; two valves control the pressure inside the TBC. A chamber is used as the vacuum reservoir and enables several activations of the TBC without the direct use of a vacuum pump; (b) Mechanical testing machine used to experimentally evaluate the TBC engaging force. In order to address the requirements of compactness and mobility in wearable devices, we developed a miniature pneumatic circuit that is managed by a custom control unit. A microcontroller (TMS320F28035 from Texas Instruments, Dallas, TX, USA) regulates the pressure inside the TBC through two pneumatic valves (TX3P030LV03LN from First Sensor AG, Berlin, Germany) and a pressure sensor (MPXV6115V from NXP Semiconductor, Eindhoven, The Netherlands). One of the two valves (valve 1 in Figure 5) connects the clutch to the ambient pressure for the disengaging mode, while the other valve (valve 2) connects the clutch to a vacuum chamber for the engaging mode. A second pressure sensor is located between valve 2 and the vacuum chamber to monitor the pressure of the chamber. A PC is connected to the setup by a serial interface (UART) to set the clutch pressure and acquire the pressure monitored by the system. A control loop is implemented in the microcontroller to maintain the set pressure. Normally, both valves are deactivated; in this condition the clutch is closed and the pressure inside is kept to the reached value. We added a hysteresis of +0.01 bar to the set value to prevent any unnecessary valve activation and waste of energy. We used a vacuum chamber as the vacuum reservoir that decouples the pump speed from the actuator speed. The vacuum chamber is brought to the desired vacuum level by the pumping system and then switched off. The developed system was used to characterize the clutch in terms of force and response time. In a real application, a constant pressure inside the reservoir can be guaranteed by the same control unit and by monitoring the pressure sensor between valve 2 and the reservoir. ator we NAS AAR RAPEEAL IE BAA Ea hy EVV VAR EEE LEE Pee ROMA SALOME UIC LR. SA LLIGAIIVEL 16 UOC Ao UIC vacuum reservoir and enables several activations of the TBC without the direct use of a vacuum pump;

Figure 5 (a) The control unit; two valves control the pressure inside the TBC. A chamber is used as the vacuum reservoir and enables several activations of the TBC without the direct use of a vacuum pump; (b) Mechanical testing machine used to experimentally evaluate the TBC engaging force. In order to address the requirements of compactness and mobility in wearable devices, we developed a miniature pneumatic circuit that is managed by a custom control unit. A microcontroller (TMS320F28035 from Texas Instruments, Dallas, TX, USA) regulates the pressure inside the TBC through two pneumatic valves (TX3P030LV03LN from First Sensor AG, Berlin, Germany) and a pressure sensor (MPXV6115V from NXP Semiconductor, Eindhoven, The Netherlands). One of the two valves (valve 1 in Figure 5) connects the clutch to the ambient pressure for the disengaging mode, while the other valve (valve 2) connects the clutch to a vacuum chamber for the engaging mode. A second pressure sensor is located between valve 2 and the vacuum chamber to monitor the pressure of the chamber. A PC is connected to the setup by a serial interface (UART) to set the clutch pressure and acquire the pressure monitored by the system. A control loop is implemented in the microcontroller to maintain the set pressure. Normally, both valves are deactivated; in this condition the clutch is closed and the pressure inside is kept to the reached value. We added a hysteresis of +0.01 bar to the set value to prevent any unnecessary valve activation and waste of energy. We used a vacuum chamber as the vacuum reservoir that decouples the pump speed from the actuator speed. The vacuum chamber is brought to the desired vacuum level by the pumping system and then switched off. The developed system was used to characterize the clutch in terms of force and response time. In a real application, a constant pressure inside the reservoir can be guaranteed by the same control unit and by monitoring the pressure sensor between valve 2 and the reservoir. ator we NAS AAR RAPEEAL IE BAA Ea hy EVV VAR EEE LEE Pee ROMA SALOME UIC LR. SA LLIGAIIVEL 16 UOC Ao UIC vacuum reservoir and enables several activations of the TBC without the direct use of a vacuum pump;

Related Figures (7)

Figure 1. (a) CAD presentation of our textile-based clutch (TBC); two layers of textile parallel to each other are placed inside an elastic cover. (b,c) Textile layers can slide in front of each other by means of an external elongation force; an elastic element in each layer of textile enables the device to elongate and recover its initial configuration. (d) Connecting the TBC to a vacuum source enables the elaborated layers to create a temporary adhesion, which blocks the elongation of the TBC under external elongation forces. each other are placed inside an elastic cover. (b,c) Textile layers can slide in front of each other by means of an external elongation force; an elastic element in each layer of textile enables the device to
Figure 1. (a) CAD presentation of our textile-based clutch (TBC); two layers of textile parallel to each other are placed inside an elastic cover. (b,c) Textile layers can slide in front of each other by means of an external elongation force; an elastic element in each layer of textile enables the device to elongate and recover its initial configuration. (d) Connecting the TBC to a vacuum source enables the elaborated layers to create a temporary adhesion, which blocks the elongation of the TBC under external elongation forces. each other are placed inside an elastic cover. (b,c) Textile layers can slide in front of each other by means of an external elongation force; an elastic element in each layer of textile enables the device to
Figure 2. (a) The bumps of one layer (blue) enter the gap between bumps of the front layer (red) and create an interlocking interaction. (b) A prototype of an array of bumps integrated on the surface of cotton webbing; the textile remains flexible after integrating the bumps. (c,d): The top and side views of one bump interlocked with two bumps in the front layer; under force of environmental pressure Fp (generated by the vacuum), each bump can withstand the engaging force of F,,., which is the sum of two forces F;, acting on the contacting points of the bumps (details in S4). Figure 2. (a) The bumps of one layer (blue) enter the gap between bumps of the front layer (red) and all “ 7’ “ Both the geometry and frictional property of the bumps can influence the result of the interlocking action during the engaging mode. In this work we selected a semispherical geometry for the bumps. During the disengage mode, bumps with a semispherical shape easily slide on top of each other and do not significantly affect the elongation force (Supplementary Materials, Video 52). In the engaging mode (vacuum applied), each single bump is subjected to a normal force generated by the pressure. This bump penetrates the front array of bumps, creating an engaging action that prevents the bumps from sliding on top of each other (Supplementary Materials, Video $3). In an array of semispherical interlocking bumps when all the bumps have penetrated each other, by assuming that the array of bumps remains planar and the bumps remain perpendicular with respect to the array plane, the interlocking force can be obtained by Equation (5). Supplementary Material 54 gives the details of Equations (5) and (6).
Figure 2. (a) The bumps of one layer (blue) enter the gap between bumps of the front layer (red) and create an interlocking interaction. (b) A prototype of an array of bumps integrated on the surface of cotton webbing; the textile remains flexible after integrating the bumps. (c,d): The top and side views of one bump interlocked with two bumps in the front layer; under force of environmental pressure Fp (generated by the vacuum), each bump can withstand the engaging force of F,,., which is the sum of two forces F;, acting on the contacting points of the bumps (details in S4). Figure 2. (a) The bumps of one layer (blue) enter the gap between bumps of the front layer (red) and all “ 7’ “ Both the geometry and frictional property of the bumps can influence the result of the interlocking action during the engaging mode. In this work we selected a semispherical geometry for the bumps. During the disengage mode, bumps with a semispherical shape easily slide on top of each other and do not significantly affect the elongation force (Supplementary Materials, Video 52). In the engaging mode (vacuum applied), each single bump is subjected to a normal force generated by the pressure. This bump penetrates the front array of bumps, creating an engaging action that prevents the bumps from sliding on top of each other (Supplementary Materials, Video $3). In an array of semispherical interlocking bumps when all the bumps have penetrated each other, by assuming that the array of bumps remains planar and the bumps remain perpendicular with respect to the array plane, the interlocking force can be obtained by Equation (5). Supplementary Material 54 gives the details of Equations (5) and (6).
2.2. PrOLOtyping We combined different techniques such as embossing, casting, and sewing to develop the TBC srototypes (Figure 3a—g). For each clutch, the array of rigid bumps was directly fabricated on a pair of voven cotton webbings (0.5 mm thickness and 30 mm width). Two series of TBC prototypes with 4mm ind 2 mm diameter semispherical bumps were made. The bumps were directly shaped and integrated yn the textile webbings by CNC milled aluminum molds (Figure 3i,j). The metallic molds enable the tray of bumps to be made quickly and accurately in one shot of pressing action. The granules of »olyoxymethylen (POM) thermoplastic with high stiffness and excellent dimensional stability [31] were irst poured into the preheated (270 °C) aluminum mold (Figure 3a,b). After fusing granules, they were oressed inside the mold’s semispherical cavities by rolling a metallic cylinder over the fused materials intil a thin, flat, and homogenous layer was obtained (Figure 3c-e and Supplementary Materials Video S1). The bumps inside the aluminum mold were transferred to the cotton webbing by locating he webbing on the top of the array and pressing it towards the melted polymer, again by rolling he metallic cylinder (Figure 3f). By pressing, the melted thermoplastic penetrates the textile porous exture and creates a strong bonding which is unified by the bump material. Finally, the assembly was emoved from the mold after cooling the mold with cold water (Figure 3g). The strong bonding of the umps to the textile guarantees the stability of the bumps in their location under shear forces generated luring the engaging action. TBCs with 4 mm bumps integrate arrays of 5 x 1 bumps in front of 4 x 11 vith a 6 mm distance between bumps (Figure 3k); TBCs with 2 mm bumps were made from arrays of 10 x 22 in front of 9 x 22 with a distance of 3 mm in between bumps ( Figure 31). The webbings vith one column less of bumps permit a symmetric configuration of webbings in the final assembly is the bumps of each webbing sit in the free space between bumps of the other webbing. An elastic extile band with 15 mm width and maximum 120% elongation (Super-E sermany) was sewn to one head of each elaborated webbing. Sewing was astic Prym Co., Stolberg selected as it is one of the yest techniques that adapts to the textile without affecting its properties or generating any rigidity in he connection point of the elastic and non-elastic bands while providing a strong bonding. The total practical length of all the TBC prototypes was 180 mm, while the length of the interfacing area was 0 mm. For all the prototypes, an elastic cover larger than the textile assembly was created by a casting srocess of silicone elastomer (Ecoflex 0030 of Smooth-On Co., Macungie, PA, USA) with a rectangular shape (50 mm xX 220 mm, with 1 mm of wall thickness). The whole assembly of textile bands was nserted and sealed inside the elastic covers (Figure 4). A 4mm silicone tube (GAINT-GOBAIN Co. -ourbevoie, France) was later integrated to the elastic cover by silicone caulk. Two other series of clutches based iust on frictional materials were created in order to compare the + + ae cal interlocking between bumps of two arrays is guaranteed. A tradeoff for the size of the bumps is required, as the larger bumps can transform the environmental pressure to a higher normal force, while smaller bumps enable a higher number of bumps to be integrated. The influence of size and density of bumps were investigated in our prototypes and experimental studies, and are reported in the next sections.
2.2. PrOLOtyping We combined different techniques such as embossing, casting, and sewing to develop the TBC srototypes (Figure 3a—g). For each clutch, the array of rigid bumps was directly fabricated on a pair of voven cotton webbings (0.5 mm thickness and 30 mm width). Two series of TBC prototypes with 4mm ind 2 mm diameter semispherical bumps were made. The bumps were directly shaped and integrated yn the textile webbings by CNC milled aluminum molds (Figure 3i,j). The metallic molds enable the tray of bumps to be made quickly and accurately in one shot of pressing action. The granules of »olyoxymethylen (POM) thermoplastic with high stiffness and excellent dimensional stability [31] were irst poured into the preheated (270 °C) aluminum mold (Figure 3a,b). After fusing granules, they were oressed inside the mold’s semispherical cavities by rolling a metallic cylinder over the fused materials intil a thin, flat, and homogenous layer was obtained (Figure 3c-e and Supplementary Materials Video S1). The bumps inside the aluminum mold were transferred to the cotton webbing by locating he webbing on the top of the array and pressing it towards the melted polymer, again by rolling he metallic cylinder (Figure 3f). By pressing, the melted thermoplastic penetrates the textile porous exture and creates a strong bonding which is unified by the bump material. Finally, the assembly was emoved from the mold after cooling the mold with cold water (Figure 3g). The strong bonding of the umps to the textile guarantees the stability of the bumps in their location under shear forces generated luring the engaging action. TBCs with 4 mm bumps integrate arrays of 5 x 1 bumps in front of 4 x 11 vith a 6 mm distance between bumps (Figure 3k); TBCs with 2 mm bumps were made from arrays of 10 x 22 in front of 9 x 22 with a distance of 3 mm in between bumps ( Figure 31). The webbings vith one column less of bumps permit a symmetric configuration of webbings in the final assembly is the bumps of each webbing sit in the free space between bumps of the other webbing. An elastic extile band with 15 mm width and maximum 120% elongation (Super-E sermany) was sewn to one head of each elaborated webbing. Sewing was astic Prym Co., Stolberg selected as it is one of the yest techniques that adapts to the textile without affecting its properties or generating any rigidity in he connection point of the elastic and non-elastic bands while providing a strong bonding. The total practical length of all the TBC prototypes was 180 mm, while the length of the interfacing area was 0 mm. For all the prototypes, an elastic cover larger than the textile assembly was created by a casting srocess of silicone elastomer (Ecoflex 0030 of Smooth-On Co., Macungie, PA, USA) with a rectangular shape (50 mm xX 220 mm, with 1 mm of wall thickness). The whole assembly of textile bands was nserted and sealed inside the elastic covers (Figure 4). A 4mm silicone tube (GAINT-GOBAIN Co. -ourbevoie, France) was later integrated to the elastic cover by silicone caulk. Two other series of clutches based iust on frictional materials were created in order to compare the + + ae cal interlocking between bumps of two arrays is guaranteed. A tradeoff for the size of the bumps is required, as the larger bumps can transform the environmental pressure to a higher normal force, while smaller bumps enable a higher number of bumps to be integrated. The influence of size and density of bumps were investigated in our prototypes and experimental studies, and are reported in the next sections.
Figure 3. Bump integration on textiles: The metallic mold is pre-heated (a) then the plastic pellets are added (b) and fused (c). The fused material is then pressed inside the mold by a roller (d) until a homogeneous flat surface is obtained (e). The textile webbing is then laid on this surface (f) and the bumps are transferred onto it (g). (k) Big bump prototype. (i) Big bump mold. (1) Small bump prototype; (j) Small bump mold. the interfacing zone; the second by coating 0.5 mm of silicone rubber at the interfacing zone, which was achieved by the direct sinking of textile webbing in silicone elastomer (Ecoflex 0030, Smooth-On, Inc.).
Figure 3. Bump integration on textiles: The metallic mold is pre-heated (a) then the plastic pellets are added (b) and fused (c). The fused material is then pressed inside the mold by a roller (d) until a homogeneous flat surface is obtained (e). The textile webbing is then laid on this surface (f) and the bumps are transferred onto it (g). (k) Big bump prototype. (i) Big bump mold. (1) Small bump prototype; (j) Small bump mold. the interfacing zone; the second by coating 0.5 mm of silicone rubber at the interfacing zone, which was achieved by the direct sinking of textile webbing in silicone elastomer (Ecoflex 0030, Smooth-On, Inc.).
Figure 4. (Top) The internal layers of the TBC with elastic bands, elaborated zone with rigid bumps, and inextensible cotton webbing; (bottom) final assembled TBC inside the air sealed silicone rubber cover.
Figure 4. (Top) The internal layers of the TBC with elastic bands, elaborated zone with rigid bumps, and inextensible cotton webbing; (bottom) final assembled TBC inside the air sealed silicone rubber cover.
Figure 6. (a) The mean maximum withstanding force of all the tested samples, and plot of Equation (5) at the default length and under different applied pressures. (b) The withstanding force of all the samples under different applied pressures and different initial elongations. (c) The force-displacement eraph of all the TBCs under —0.2 atm, and (d) under —0.8 vacuum pressure.
Figure 6. (a) The mean maximum withstanding force of all the tested samples, and plot of Equation (5) at the default length and under different applied pressures. (b) The withstanding force of all the samples under different applied pressures and different initial elongations. (c) The force-displacement eraph of all the TBCs under —0.2 atm, and (d) under —0.8 vacuum pressure.
Figure 7. (a) The engaging response time of TBCs with small bumps. (b) The engaging response time of TBCs with large bumps (standard deviation +3-8 ms). Both types of samples were tested in two different conditions; one based on ambient pressure and one based on a pre-charged pressure of —0.01 atm.
Figure 7. (a) The engaging response time of TBCs with small bumps. (b) The engaging response time of TBCs with large bumps (standard deviation +3-8 ms). Both types of samples were tested in two different conditions; one based on ambient pressure and one based on a pre-charged pressure of —0.01 atm.