A Study on the Effects of Graphite Flakes on Torque Transmission of a Magnetorheological Fluid Clutch

Abstract

The transmittable torque of magnetorheological fluid (MRF) clutches can be adjusted by controlling the magnetic field inside the clutch. In order to increase their maximum transmittable torque while decreasing their volume and mass, geometry optimization is often carried out, such as in [1]. Additionally, the MRF used in them can be designed to have a higher yield stress, resulting in a higher transmittable torque. In [2], this was studied by adding micrometer-sized graphite flakes to an in-house created MRF; it was found that the inclusion of graphite flakes led to an increased yield stress. The present research aims to further study this, by investigating the effects that including graphite flakes in a commercially available MRF may have on the transmittable torque of an MRF clutch. Clutch design. The MRF clutch utilized in this research is a drum-type, permanent magnet-based clutch, made from two concentric cylinders. The external cylinder was fabricated from low carbon steel and the internal cylinder was fabricated from aluminum. A cavity for the MRF is formed between these cylinders, with a width of 0.75 mm. The external cylinder formed the main body of the output stage of the MRF clutch. The magnet assembly used to magnetize the clutch consisted of four magnets which have a hollowed-out cylinder shape when assembled together. These magnets were magnetized in the radial direction, with two of them having their south poles in the outer face, while the other two had them in the inner face. The internal cylinder formed the main body of the input stage of the clutch, driven by a DC motor. A mechanism called field blocker, presented in [3], was used to adjust the magnetization level of the MRF inside the clutch. MRF. A commercially available MRF, the LORD MRF-140CG, was used in this study. Four mixtures with 20 μm graphite flakes (282863, Sigma-Aldrich) were prepared, with a graphite content of 0.5% w/w, 1.0% w/w, 2.0% w/w, and 3.0% w/w. Experimental set-up. The MRF clutch was driven by a velocity-controlled DC motor, and its output stage was connected to one end of a 5 Nm T22 HBM torque meter. A lever was attached to the other end of this torque meter, which contacted the testing bench when rotated in one or the other direction. After contacting the bench, this lever forced the output stage of the MRF clutch to remain stationary while the input stage continued rotating. A stepper motor moved the field blocker in or out of the MRF clutch by means of a ball-screw mechanism, covering and uncovering the magnet assembly. The testing set-up is shown in Figure 1 a). Experimental procedure. The testing of the four MRF mixtures together with the MRF without graphite were carried out the following way: 1) The torque output was measured during one second without driving the clutch and with the lever in a free position. 2) The clutch was then driven at the velocity under test for a duration equal to 2 s plus the time required for the clutch to complete a rotation. 3) The clutch was driven backward at a low angular velocity for one second, in order to leave the lever in a free position. This procedure was repeated three times for each combination of five angular velocities, six field blocker positions, with and without the magnet assembly, for a total of 180 measurements per MRF mixture. The average torque value was calculated from the torque recorded during the second testing step, after the first two seconds had passed. The effects on torque due to seal friction and viscosity on the measured torque values were then removed. Results. The results of these tests are summarized in Table 1, Figure 1 b) and c). In Figure 1 b) and c), the average torque values of the tests carried out with the field blocker not covering and covering the magnets are shown (two out of the six positions tested). The torque values are higher in the case were the magnet assembly is uncovered than when it is covered. It is also seen that the difference in torque caused by the graphite content increases as the input angular velocity increases. Higher magnetization levels decrease the difference in torque (the field blocker does not block the magnetic field entirely). The mean, maximum, and minimum torque values over the six magnetization cases are shown in Table 1. It can be seen that on average, the transmittable torque increases with both graphite content and angular velocity, achieving up to 70% torque increase. Conclusion. The inclusion of graphite flakes in a commercially available MRF can be used to increase the output torque of an MRF clutch, a result which aligns to those of [2]. This effect in a clutch with aluminum cylinders is more evident at higher angular velocities. **