2. Mechanism Design Procedure

To design a linear drive system to the robotic gripper that will have both a slow close and quick snap open, the team first setup their design constraints. The design constraints of the project are shown below.

Design Constraints

• Snapping Motion without a ramp up period
• No pneumatics
• Single motor control
  • Spin CCW to close
  • Spin CW to snap open
• Provide user control on the exact time to release
• Instant sliding block velocity of 1 m/s when the mechanism releases

The listed constraints above were formulated internally amongst the team. The snapping motion without a ramp up period was set by the team because if you were to try and toss a water bottle with your fingers, those fingers would actually need to release almost instantaneously at the right moment to match the swinging motion of your arm with the opening motion of you thumb and index finger let go of the water bottle. This constraint therefore eliminated a quick release linkage system, shown in Figure 1. This type of mechanism does release quickly, but not until the link being driving by the wheel with a dowl is aligned directly above the drive wheel's center of rotation.

Figure 1: Quick Release Linkage Mechanism

The next constraint limiting the use of no pneumatics was set based on the fact that it is not common for many electro-mechanical robotic arms to be equipped with a supply of compressed air nor the hosing need to supply this air at the end of the robotic arm. Pneumatics are great for a snapping motion, but without a supply of compressed gas you are going nowhere.

Single motor control was another constraint that the team limited themselves to because of two things, one in weight of the overall drive mechanism and the other is control. Typically, a robotic arm does not have extensive strength for the amount of weight they can manipulate and move around. By adding a heavy end effector to the end of the arm, you will be decreasing the amount of weight the arm has left over to manipulate objects in the world. The second reason for a single motor is control. robotic arms generally only have one power supply at the end of the robotic arm for just the addition of a single motor. Plus having one motor means the team can provide easier control to the user as they only need to control the direction of the motor. The intent of the design is to have the user just need to swap the direction of the spin on the motor itself and the gripper will act accordingly.

Providing a user the ability to release the water bottle at the right moment is key, as mentioned earlier. Without the ability to time the snap open correctly, flipping a water bottle would be almost impossible. Using a typical linkage mechanism quickly release the linkage system would be very challenging, as the point as to where the motion turns out to snap would need software that knows the motors position, which would need to be based on encoders from a motor, which are not always accurate. Therefore a software based release procedure would not be robust and could lead to the user dropping the water bottle at the incorrect moment.

Instantaneous sliding block velocity of 1 m/s was derived based on [1]. In this paper the authors calculated the angular velocity your finger moves with respect to your thumb when you snap your fingers. This value was found to be on average roughly 24 rad/s, for an average human being. Therefore, the team did back calculations on a hypothetical robotic gripper to get a linear velocity on the drive rail from requiring the fingers on the hypothetical gripper to open at 24 rad/s. The linear velocity at the rail was found to be 1 m/s. This velocity was then used in our overall design.

Design Iterations

With the design constraints set, the team went to work brainstorming many ideas. The first idea that the team landed on is shown in Figure 2. This design idea was to use a four-bar linkage to actuate the linear bar. The problem with this first design was that the linkages would never be able to disconnect from the drive motor without an extra mechanism, but provided the linear actuation part of the mechanism with low friction. 

Figure 2: Design Idea One. Perform linear translation with a four-bar linkage.

To try and design a simple method to decouple the drive motor, since they have immense friction when trying to back drive. The next idea the team had is shown in Figure 3. This drawing shows an idea where a drive motor would drive a swivel block that was on a lead screw. The swivel block has a lever that would drive the actual rail the robotic gripper was attached to when spinning in one direction. Then when the user want to release the user would spin the motor the opposite direction to quickly release the spring loaded block. This mechanism would work in concept, but with so much force on the spring loaded block, the lever action to turn the drive block out of the way would never open with just spinning the motor the opposite direction.

Figure 3: Design Idea Two. Create a linear translation with a lead screw and sprung shaft that unlatches to release the spring-loaded block.

Another idea that the team had to drive this linear actuation rail connected to the gripper is shown in Figure 4. This drive mechanism was based off a rack and pinion design where a drive motor and gear would be connected to the spring-loaded shaft that drives the entire assembly to close. When the user was ready to release they would be able to disconnect the gear from the rack and spring back. The issue with this design is that the motor and everything would be moving as one, which add more weight that the spring had to force backwards. It also did not allow the user to drive the motor CCW or CW to close and release. Also with this an issue arises with gear meshing when trying to have the pinion gear engage with the rack.

Figure 4: Design Idea Three. Rack and pinion design for opening and closing the gripper

After some thought, the team decided to stick with their Figure 2 design, but with a modified version to free the drive motor from the spring loaded actuator. In Figure 5 below, a disc mechanism is shown, which allows to connect the two dot line links with the outer disc when rotating one way, then to disconnect the links when spinning the other direction. A better version of this Figure 5 and Figure 1 were combined to produce the final prototype of our mechanism.

Figure 5: Early Idea of Drive Mechanism that allows for the drive mechanism to decouple from the spring-loaded linear actuator

Final Design

The final design that was decided upon is shown below in Figure 6. The mechanism shown below in Figure 6, is a three part mechanism. The green spindles are similar to a drum brake design where when they spin out they catch the top lid of the mechanism and make both the spindles and the top lid spin together as one. Then the user wants to release the motor is spun in the opposite direction where the ratchet gear catches the black gear shown and allows for the green spindle to spin in and release the top lid which is connected to the lid.

Figure 6: Decoupling mechanism used to decouple the drive motor when you want to release the mechanism and then connect the motor when you want drive the motor.

References

[1] R. Acharya, E. J. Challita, M. Ilton, and M. S. Bhamla, “The ultrafast snap of a finger is mediated by skin friction,” p. 12.