Gripping characteristics of an electromagnetically activated magnetorheological fluid-based gripper

The design and test of a magnetorheological fluid (MRF)-based universal gripper (MR gripper) are presented in this study. The MR gripper was developed to have a simple design, but with the ability to produce reliable gripping and handling of a wide range of simple objects. The MR gripper design consists of a bladder mounted atop an electromagnet, where the bladder is filled with an MRF, which was formulated to have long-term stable sedimentation stability, that was synthesized using a high viscosity linear polysiloxane (HVLP) carrier fluid with a carbonyl iron particle (CIP) volume fraction of 35%. Two bladders were fabricated: a magnetizable bladder using a magnetorheological elastomer (MRE), and a passive (non-magnetizable) silicone rubber bladder. The holding force and applied (initial compression) force of the MR gripper for a bladder fill volume of 75% were experimentally measured, for both magnetizable and passive bladders, using a servohydraulic material testing machine for a range of objects. The gripping performance of the MR gripper using an MRE bladder was compared to that of the MR gripper using a passive bladder.


I. INTRODUCTION
Robotic grippers have been developed to emulate the dexterity of the human hand because the human hand can perform not only power grips, but also precision grips for delicate and precise tasks in various work scenarios.As a result, the design of robotic grippers have become increasingly more complex, requiring extensive sensors, actuators, and controllers to enable their implementation.In addition, sophisticated control algorithms with a priori knowledge of an object's physical properties such as shape, size, and texture, etc. are often required for even a simple gripping task. 1,2The increasing complexity of these systems creates a higher risk of system failure as well as increased mass, cost and power requirements.
3][4] Granular-jamming universal grippers have an elastic bladder that is filled with a granular material and passively conforms to an object's shape and size.Then, to grip the object, interstitial air is quickly evacuated from the elastic bladder using a vacuum pump, thereby inducing granular jamming, so that gripping force can be exerted on the object.Universal granular-jamming grippers are simple and low cost but can grip a wide range of objects without relying on extensive sensors, actuators and feedback control.However, the universal granular-jamming gripper requires a vacuum pump and an air source to evacuate or inflate the bladder to enable gipping and release of objects, which limits how compact the gripping system due to the added volume and weight of the air handling components.In addition, since the a Corresponding author.Electronic mail: wereley@umd.edu.gripping force of granular jamming grippers depends on the pressure difference between the inside and outside of the elastic bladder, it may not be practical for environments such underwater and space having pressure and temperature extremes. 4ore recently, universal grippers have been developed using magnetorheological fluids (MRFs) having rheological properties, such as yield stress and viscosity, that can be controlled using a magnetic field.Such MRF-based grippers (MR gripper) eliminate the need for on-board bulky, heavy, and noisy vacuum pumps used in granular jamming grippers, 5,6 enabling lighter and more compact designs.To grip an object, the bladder is pressed around the object, and the electromagnet is activated, so that the MRF forms particle chains and hardens around the object, thereby enabling the gripping action.Prior MRF-based grippers used a passive bladder.In this study, an MRF-based universal gripper with a magnetizable magnetorheological elastomer (MRE) bladder was developed having a simple configuration designed to achieve larger holding force and more reliable gripping for a wide range of objects, and its effectiveness is experimentally evaluated and compared to an MRF-based gripper with a passive bladder.

II. DESIGN OF THE MR GRIPPER
Figure 1 presents a photo and schematic of the MR gripper.As seen in this figure, the MR gripper has a simple design in which a cup-type rubber bladder filled with MRF was mounted atop the electromagnet (4.25" in diameter and 1.5" in height, with 7.8 Ω in electrical resistance).The rubber bladder is the compliant component of this system and is elastically deformed by gripped objects when pressed onto gripped objects.In this study, two different rubber bladders were fabricated and tested: one is non-magnetizable (passive) rubber bladder and the other is magnetizable MRE bladder.The passive rubber bladder was made by curing liquid silicone rubber (shore 00-50) in a 3D printed plastic mold.The MRE bladder was made by dispersing magnetizable iron oxide particles (25% in volume fraction, hereinafter vol%) into the liquid silicone rubber (shore 00-20) and cured in the same mold.It should be noted that the elastic stiffness of the passive rubber bladder was matched to that of the MRE bladder so as to produce similar bulge stiffness.The MRF synthesized for this study used a high viscosity linear polysiloxane (HVLP) carrier fluid (viscosity of 20,000 cSt), in which were dispersed carbonyl iron particles (35 vol%).This MRF has been shown to have excellent long-term sedimentation stability. 7Here, 75% of the bladder volume was filled with MRF.When the bladder is pressed around an object, and the electromagnet is activated using applied current input, the MRF contained within the rubber bladder hardens due to particle chain formations induced by the applied electromagnetic field, thereby locking up the bladder around the object enabling it to grip the object.The MRE bladder can increase this lock-up effect of the MRF because the MRE bladder itself is magnetizable and thus moves toward the electromagnet when energized.On the other hand, the center bobbin worked as the magnetic core of the electromagnet.In this study, the depth between the top surface of the electromagnetic coil and the top surface of the magnetic core was adjustable by changing the height of the magnetic core inserted inside the electromagnetic coil.As a result, the gripped objects could be more encompassed by more effective magnetic field so as to improve the holding performance of the MR gripper.

III. EXPERIMENTAL TESTING
Figure 2 presents the experimental test setup used to measure gripping performance of the MR grippers.The top end of the hydraulic testing machine (Instron Dynamight 8841) was connected to the MR gripper and the target object was fixed to the load cell.A power supply was connected to the MR gripper to provide current to the electromagnet.Firstly, the target object was pressed into the MR gripper with an applied (initial compression) force.The bladder of the MR gripper deformed to partially envelope the target object.A current was then supplied to the MR gripper to grip the object.The holding force of the MR gripper was measured from the load cell when the MR gripper was vertically lifted up by the hydraulic actuator with an extension force increment rate of 2 N/s.Three different diameters of plastic cylinders (1", 1.5", and 1.75" in diameter) and two different sizes of wood spheres (1" and 1.5" in diameter) were chosen as the target objects to be gripped.In addition, cylinders were investigated that had either a smooth surface or a surface with vertically oriented ridges.

IV. RESULTS AND DISCUSSION
Figure 3 shows measured holding force vs. displacement of the MR gripper for a 1.75" cylinder and an applied current of 3 A. As displacement increased, the holding force also increased.When the MR gripper became detached from the object to be grasped, the holding force dropped toward zero.The holding force of the MRE bladder case rapidly became zero after a peak value as the displacement increased.But, the holding force of the passive bladder case was smoothly reduced after a peak value and finally became zero after much larger displacement than the MRE bladder case.For an applied force of 40 N, the peak holding force for the MRE bladder (56.9 N) was much greater than that for the passive bladder (35.1 N), or increase in gripping performance of 62%.
Figure 4 presents measured peak holding force versus applied force for the tested objects at a constant current input of 3 A. In this case, the holding force measurement was conducted five times and the average values reported in the figure.From this data, the holding force of the MR gripper increased as the applied force increased.The larger applied force forced the object to be more enveloped by the bladder and to bring it closer to the electromagnet.Depending on the object, the holding force of the MR gripper dramatically changed.For the 1" diameter objects shown in Figure 4(a), the holding force of both sphere and cylinder were quite similar.But, for the 1.5" diameter objects shown in Figure 4(b), the holding force for the cylinder was significantly larger than that for the sphere.Thus, the object shape greatly contributed to holding performance, since the MR gripper tended to hold cylinders better than spheres for objects greater than 1" in diameter.For 1.75" cylinders as shown in Figure 4(c), the cylinder with a smooth surface showed much larger holding force than the vertically FIG. 2. Experimental test setup to evaluate gripping performance of the MRF-based universal gripper with the MRE bladder.ridged (or serrated) surface case, because the ridges reduced the contact area between the bladder and the perimeter of the object as Figure 4(d), had a large impact holding force: a larger size object produced a larger holding force.The gripping action of the MR gripper results from the friction force of the contact surface between the bladder and the gripped object, which increases for larger objects.On the other hand, as shown in Figure 4(d), the MR gripper with an MRE bladder showed better holding force performance than the MR gripper with a passive bladder.For the MRE bladder, the additional attractive magnetic force of the MRE bladder itself helped to improve the gripping action.The largest average peak holding force for the MRE bladder  5(d) for all smooth-walled cylinders) gripped using both passive and MRE In this case, the holding performance index was defined by the ratio of the peak holding force to the applied force.A holding performance index ≥ 1 implies that the holding force of the MR gripper is equal to or greater than the applied force to the object.As observed in this figure, the holding performance index mostly decreased as the applied force increased.The MR gripper showed larger holding performance index at the lower applied force (below about 20 N) range than the MR gripper.But, in higher applied force range, the MR gripper with an MRE bladder produced the best holding performance index.

CONCLUSIONS
In this study, testing of a novel MR gripper design capable of grasping a wide range of objects was presented.The MR gripper fabricated here had a simple configuration in which a cup-type elastic bladder filled (75% fill volume) with a magnetorheological fluid (MRF), synthesized by dispersing 35 vol% carbonyl iron particles in a highly viscous linear polysiloxane (HVLP) carrier fluid, and mounted atop a round-type electromagnet.Two different MR grippers were proposed and fabricated: one used a passive silicone rubber bladder the other used a magnetizable magnetorheological elastomer (MRE) bladder.MR grippers utilizing either a passive or an MRE bladder were able to grip a group of objects including 1", 1.5" and 1.75" diameter cylinders or spheres.Both 1" diameter objects were gripped with similar holding force using either MR gripper configuration.The MR gripper using the MRE bladder generally produced larger holding forces than that using the passive bladder for 1.5" 1.75" cylinders, although for 1.75" cylinders, this was only true for applied forces greater than 25 N. Thus, the application of MRE bladders in MR gripper systems has been shown to be feasible and to have performance benefits, and further study is warranted to further examine performance, durability, and effect of volume fraction of carbonyl iron powder fill in the MRE bladder material.

FIG. 3 .
FIG.3.Measured holding force of the MRF-based universal grippers versus displacement when a target object (1.75" cylinder) was initially gripped with an applied (compression) force of 40 N.